Biosimilars - University of Delhimicrobio.du.ac.in/web3/uploads/Microbiology Uploads... · Biosimilars: A Philosophy? Setting the Scene for Biosimilars Biological medicinal products

Biosimilars

Jean-Louis PrugnaudJean-Hugues Trouvin

Editors

Biosimilars

A New Generation of Biologics

123

EditorsJean-Louis PrugnaudHôpital Saint-AntoineParisFrance

Jean-Hugues TrouvinUniversité Paris DescartesParisFrance

ISBN 978-2-8178-0335-7 ISBN 978-2-8178-0336-4 (eBook)DOI 10.1007/978-2-8178-0336-4Springer Paris Heidelberg New York Dordrecht London

Library of Congress Control Number: 2012949364

� Springer-Verlag France 2013This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part ofthe material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission orinformation storage and retrieval, electronic adaptation, computer software, or by similar or dissimilarmethodology now known or hereafter developed. Exempted from this legal reservation are briefexcerpts in connection with reviews or scholarly analysis or material supplied specifically for thepurpose of being entered and executed on a computer system, for exclusive use by the purchaser of thework. Duplication of this publication or parts thereof is permitted only under the provisions ofthe Copyright Law of the Publisher’s location, in its current version, and permission for use must alwaysbe obtained from Springer. Permissions for use may be obtained through RightsLink at the CopyrightClearance Center. Violations are liable to prosecution under the respective Copyright Law.The use of general descriptive names, registered names, trademarks, service marks, etc. in thispublication does not imply, even in the absence of a specific statement, that such names are exemptfrom the relevant protective laws and regulations and therefore free for general use.While the advice and information in this book are believed to be true and accurate at the date ofpublication, neither the authors nor the editors nor the publisher can accept any legal responsibility forany errors or omissions that may be made. The publisher makes no warranty, express or implied, withrespect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Translation from the French language edition ‘Les biosimilaires’ by Jean-Louis Prugnaud, Jean-HuguesTrouvin (Eds.), � Springer-Verlag France, Paris, 2011; ISBN: 978-2-8178-0036-3Translated by: Mrs. Laurence Grenier

Foreword

Biosimilars: A Philosophy?

Setting the Scene for Biosimilars

Biological medicinal products including biotechnology-derived medicinal prod-ucts (often referred to as ‘‘biologicals’’) have an impressive record in treatingnumerous serious diseases, and their market is growing faster than that of allpharmaceuticals combined. Insulin produced by using recombinant DNA tech-nology was the first approved therapeutic protein. It entered the US market inOctober 1982 and subsequently also gained its marketing authorisation in Europe.Since that time, which created, a ‘‘gold rush mood’’ of sorts, the progress in theresearch and development of these innovative medicinal products has acceleratedsignificantly. At present, several hundreds of biologicals (in the broader sense)have been approved in Europe and the United States, and the number of appli-cations for marketing authorisation is still rising. However, the research anddevelopment on biotechnology-derived medicinal products is costly, includingalso the considerable work one needs to invest into defining and maintaining awell-controlled manufacturing process. Therefore, high costs must often be paidwhen it comes to treating patients, which represents a burden to health care sys-tems, and thus might limit access of patients to these medicines. The upcomingexpiration of patents and/or data protection for the first innovative biotherapeuticshas apparently created another ‘‘gold rush mood’’ to work on ‘‘generic versions’’of products ‘‘similar’’ to the originals and relying in part for their licencing on datafrom these originator products for their licencing. As we will learn in this book,however, the term ‘‘generic’’ cannot be used for such biologicals that are ‘‘copyversions’’ of licenced products. The most important reason is that biologicals arecomplex both as regards their structure and their manufacturing process. Even withvery sensitive state-of-the-art physicochemical and biological characterisationmethods, one cannot conclude that an originator product and a copy version of itare ‘‘essentially similar’’ or even ‘‘identical’’. There could be minute differences,and these differences could have a huge impact on nonclinical and clinicalbehaviour, such as safety or efficacy. This is why the term ‘‘biogeneric’’ isobsolete, and another term, namely ‘‘similar biological medicinal product’’, or, inwidely used jargon, ‘‘biosimilars’’, has been coined. Licencing of biosimilars has

v

already become a reality in Europe, since the European Medicines Agency’sscientific Committee for Medicinal Products for Human Use (CHMP) has estab-lished a regulatory framework for similar biological medicinal products for theirfacilitated development, and issued a positive ruling on two ‘‘biosimilar’’ appli-cations in 2006 (Omni trope and Valtropin, both of which are growth hormones).Many aspects of this framework and its application will be discussed in this book.Although the licencing of the first biosimilars was an initial step for other productsfollowing the biosimilar route, such an approach also has its limitations in thatclinical trials are still required, so no real ‘‘generic’’ route is yet possible. Somegeneral aspects are, however, already clear now: (1) guidelines for the develop-ment of biosimilar products can be implemented, (2) many aspects from ‘‘com-parability exercise’’ that regulators are already well accustomed with from changesin manufacturing processes of biologicals are applicable and have been readilyapplied already, (3) as shown in the case of Valtropin, the use of different hostcells for the biosimilar product and the comparator in principle can be possible,and (4) although required, the clinical programme might be abridged as comparedto a full development and has the primary aim of establishing ‘‘similarity’’. Themain principles for the development of a biosimilar medicinal product in theEuropean Union currently include:

• Demonstration of ‘‘biosimilarity’’ in terms of quality, safety and efficacy to areference product that is licenced in the EU. Therefore, all main studies shouldbe strictly comparative, and the same reference product should be usedthroughout the development programme.

• Use of a ‘sensitive’ test model (and here patients are also included, as discussedfurther below), able to detect potential differences between the biosimilar andthe reference product.

• Demonstration of equivalent rather than non-inferior efficacy, as the latter doesnot exclude the possibility of superior efficacy (only an ‘‘equivalence’’ trial cannormally a priori, on formal grounds, establish ‘‘similarity’’).

This last point is a frequently asked question: What if the biosimilar is indeedmore efficacious than the reference medicinal product? Would that not be desirablefor the patient? However, the answer is clear: a superior efficacy would not onlycontradict the assumption of similarity, and thus potentially preclude extrapolationto other indications of the reference product, particularly those with different doserequirements; it could also imply safety concerns, since in case of higher potencythe product could also be more ‘‘efficacious’’ in an unwanted sense, and safetyissues could arise when using the dose(s) recommended for the reference product,especially for drugs with a narrow therapeutic index. If similarity with the refer-ence product has been convincingly demonstrated in a key indication, extrapola-tion of efficacy and safety data to other indication(s) of the reference product, notstudied during development, may be possible if scientifically justified. This issometimes not straight forward, since it usually requires several prerequisites likeinvolvement of the same mechanism(s) of action for each of the indications, or for

vi Foreword

example the involvement of the same receptor(s) for mediating this mechanism(s)of action. This is one of the aspects that will also be discussed later in this book. Itis clear that, although a reduction in the data package is possible for a biosimilar,the pre-licencing data package is still substantial. The dossier for the CMC part(Chemistry, Manufacturing, Controls, or also the ‘‘quality’’ dossier in regulatoryjargon) has even higher requirements than a ‘‘stand-alone’’ development of a newbiological drug: First, a full quality dossier has to be provided that meets regu-latory standards to the same level as a novel biological medicinal product. Second,on top, a comprehensive comparability exercise to the reference product on thedrug substance and drug product level is required, which is an additionalrequirement, exclusive to a biosimilars. Likewise, human efficacy and safety dataare always required, but—and this is at the heart of a biosimilar development—to alesser extent than for the development of a novel medicinal product. Here, and inthe possibility to omit certain non-clinical studies usually required for a novelcompound, lies the possibility to reduce the costs of development considerably.The amount of possible reduction in non-clinical and clinical data requirementdepends on various considerations, including how well the molecule can becharacterised by state-of-the-art analytical methods, on observed differencesbetween the biosimilar and the reference product, and on the clinical experiencegained with the reference product and/or the substance class in general.

Biosimilarity: A Philosophy?

The difficulty for the developers of any biosimilar is that there is usually no directaccess to originator companies’ proprietary data. The developer of a biosimilar,thus, has to purchase the reference medicine from a pharmacy and then purify thedrug substance, and engineer a process to produce the biosimilar—that is, thedevelopment of a biosimilar requires the establishment of a new manufacturingprocess ‘‘from scratch’’. If the European biosimilar framework had required anidentical manufacturing process, then this would have automatically made bio-similar development programmes difficult if not impossible. It is acknowledged inthe respective biosimilar guide-line (Guideline on Similar Biological MedicinalProducts Containing Biotechnology-Derived Proteins as Active Substance:Quality Issues CHMP/49348/05) that ‘‘it is not expected that the quality attributesin the similar biological and reference medicinal products will be identical’’. Thisfollows from the principle that a biotechnological product is defined by the way itis manufactured, including all process-related and product related impurities,microheterogeneities, excipients, etc (‘‘process determines product’’, or ‘‘theprocess is the product’’). Due to the complex production method of biologicalmedicines, the active substance may differ slightly between the biological refer-ence and the biosimilar medicine—what ‘‘slightly’’ will mean, however, willunfortunately be a case-by-case decision based on data and involving variousconsiderations like the complexity of the molecule in question, the inherent

Foreword vii

variability known from the reference medicinal product, the potential clinicalimpact etc. One can easily recognise that a biological is therefore more than justthe active substance, and includes the aforementioned impurities etc. ‘‘Slight’’differences can have a major impact, but, theoretically, conversely some differ-ences e.g. in impurities may have no impact at all. So how to solve this problem toestablish biosimilarity? First of all, there must be a backbone of extensive andstate-of-the-art data from chemistry, manufacturing and control (‘‘quality data’’)that not only satisfy pharmacopoeias, but is also strictly comparative to the ref-erence product. This serves as the basis for abridging the non-clinical and clinicaldata requirements. When it comes to clinical data, there is at least one decisivedifference to clinical development programmes for ‘‘stand-alone’’ developmentslike biologicals with a novel mechanism of action, and this difference in concept issometimes hard for clinicians to swallow: The aim of a biosimilarity developmentprogramme is not to establish benefits for the patient—this has been establishedalready by the reference product years ago. The aim of a biosimilarity develop-ment programme is to establish biosimilarity and if there are clinically relevantdifferences, by employing a clinically relevant model. Indeed, patients are seenhere, formally speaking, as a ‘‘model’’ to establish biosimilarity. This means thatthe trial design including the primary endpoint, secondary endpoints, choice ofpatients etc may follow a different philosophy than for a novel compound. Forexample, for a clinical indication that can present itself in different severities, itmay be unwise to include patients suffering from different grades of the disease. Ifthere is—despite randomization and/or stratification—an uneven distribution ofthese numerous factors like different disease history, different pre-treatments,different disease presentation at study entry etc, then differences between the twostudy arms comparing the reference biological medicinal product with the bio-similar may be difficult to interpret: Is a perceived difference related to differencesin the molecules? If so, what if there are no measurable differences on an ana-lytical level? Or could differences be explained by differences in the patientsbetween the groups? In other words—by not focusing on a homogeneous patientpopulation, it could well be that the endpoint measures differences in diseasemanifestation and not differences between the molecules. Likewise, following thisphilosophy, the most sensitive clinical end-point may be more suitable than anendpoint establishing clinical benefit. Currently, anticancer biologicals utilising acytotoxic mechanism of action are not yet approved as biosimilars; however,upcoming discussions will have to elucidate on what endpoint to choose for suchscenarios—should it be a more sensitive and measurable end-point, e.g. tumourresponse rate, or should it be a more clinically relevant endpoint like overallsurvival of cancer patients? The tumour response rate only measures the action ofa drug, not patients’ benefit. A highly active compound that yields in a hightumour response rate could, at the same time, be considerably toxic, and thusreduce survival, which would obviously not be a benefit for patients whilst stillmeeting the endpoint ‘‘tumour response rate’’. Therefore, for new compounds, atime-related benefit endpoint is usually required, e.g. overall survival. However,one could argue that the survival benefit has already been established years before

viii Foreword

by the originator product, and that for the biosimilar this does not have to berepeated. Such considerations are currently hotly debated. Now that biosimilarshave become a reality, developers are extending their reach to more complexmolecules—including monoclonal antibodies that are much more complex thancurrently licenced biosimilars (e.g., growth hormones). It is, therefore, time todiscuss the current state-of-the-art, condensing all relevant aspects within a singlebook. I am sure that this book will not only be useful for developers of biosimilarsor for regulators—I do think that also physicians, who are the ‘‘users’’ of bio-similars, should know about how biosimilars are designed and developed, sincethey are clearly different from generics, which surely has clinical implications.

Christian K. SchneiderCHMP Working Party on Similar Biological (Biosimilar)

Medicinal Products Working Party (BMWP), European MedicinesAgency, London, United Kingdom

Paul-Ehrlich-Institut,Federal Agency for Sera and Vaccines, Langen, Germany

Twincore Centre for Experimental and Clinical Infection Research

Foreword ix

Preface

The biological medicinal products’ market has considerably expanded since 1998.The worldwide sales for these types of medicines have increased much faster thanother types of medicines—12 % on average between 1998 and 2007 versus only 4% for the sector beside biological medicinal products. The share of these in theglobal market will have risen from 10 to 15 % between 2007 and 2012, accordingto IMS (Intercontinental Marketing1 lists 633 biological medicinal products arebeing developed worldwide to treat more than 100 diseases; including: 254 drugsdeveloped for cancer; 162 for infectious diseases; 59 for auto-immune diseases and34 related to HIV/Aids pathologies).

Numerous recombinant proteins are currently in the public domain afterexpiration of the patents that protected them, thus they are an interesting target forclassic generics companies. If the pressure put on by institutions that providepayment services and a simpler licencing process have contributed to their verylarge development, then the difficulty to develop copies of biotechnologicalproducts could be a factor of weaker progression.

The term ‘‘generic medicinal product’’ is used to describe a medicine that hasan active substance made of a small, chemically synthesised molecule with a well-known structure and a therapeutic action equivalent to the original product’s.Generally, the demonstration of bioequivalence with a comparator through bio-availability studies is enough to deduct the therapeutic equivalence between thegeneric and reference medicine. This approach is not considered sufficient for thedevelopment, evaluation and approval of a biological medicine claiming itssimilarity to a reference medicine because of the molecular complexity and thedifficulty to characterise active structures. On top of that, efficacy and therapeuticsafety may be influenced by the biological source and the manufacturing process.Clinical studies are therefore necessary to demonstrate the efficacy and safety ofthese copies. As these copies are not identical to their originator but only ‘‘simi-lar’’, they are called ‘‘biosimilars’’, as a contraction of the official European des-ignation of ‘‘biological medicinal product similar to a reference biologicalmedicinal product.’’ Other designations can be found in the literature as ‘‘bioge-nerics’’ but this term can not be retained because of the ‘‘only similar’’ feature that

1 IMS Health analyse Développement and Conseil, juillet 1998.

xi

a copy may have. In the U.S.A., the term follow-on biological product (FOBP) isused to designate the copies of bio medicinal products. The World Health Orga-nisation (WHO) uses the term similar biotherapeutic product (SBP) to designatebiosimilars.2

The purpose of this book is to show how biosimilars are developed, what thecriteria and aspects that are taken into account for their licencing are, how patientssafety is preserved, what it is about the particular angle of immunogenicity, whatresponse must be considered concerning substitution and interchangeability ofthese products, what particular follow-up must be implemented (in terms ofpharmacovigilance and traceability) and what the perceptions of the players,prescribers and dispensers of these products are. Biosimilars are medicines des-tined to be present in doctors’ therapeutic tool box. Then, this book tackles thecertain aspects of strategies underlying the use of biosimilars and the resultingmedical responsibility, if this latter may ever be particular for this new type ofmedicine.

This book limits itself to the analysis of European licencing of biosimilars as itis currently on a worldwide basis, the most advanced regulation since the first 2001guidelines that allow with time to build constructive supports designed to helpindustrialists or companies to develop biosimilars.

The marketing of biosimilars is characterised by many more barriers than forgenerics; that is to say, developments necessary in order to master their manu-facturing and their quality, safety and efficacy evaluation. The biosimilars markethas induced high development costs, and higher risks; it needs a longer devel-opment time and an expertise related to the clinical development of these products.Biosimilars development strategies are not the same as those of generics. Theirdevelopment complexity as well as their production costs will favour companieswith significant financial resources, experience in the field of biological medicinalproducts production and even an expertise in marketing innovating products.

In relation to biological medicinal products development costs, the average costof treatment per patient for this category of products is much higher. Worldwide, 7medicines of the ‘‘top 10’’ sales will be in 2014 products of biological origin andtheir individual cost will comprise between 10,000 and 100,000 Euros per person.The appeal of the marketing of copies similar to the original product ensuring thesame level of quality, safety and efficacy is obvious for the organisations thatprovide payment services, as well the as the patients with incomplete or no cov-erage. However, because of the complexity of the development and production ofbiosimilars, a reflection on the necessary reduction of these products’ costs isneeded. Will this reduction be as significant as for generics?

To the effect of the cost less attractive for biosimilar prescriptions may be addeda stronger reluctance to use biological products (in their vast majority) that only

2 Medicines in development. Biotechnology. Billy Tauzin. 2008Guidelines on Evaluation ofSimilar Biotherapeutic Products (SBPs). WHO/BS/09.2110. Source: EP Vantage June 2009;Evaluate Pharma: World preview 2014, report.

xii Preface

specialists could master. The statute of prescription and dispensation of biologicalmedicinal products is subject to international regulations based upon internationalstudies that often underline the complexity of their use and of patient follow-up.Biosimilars do not escape from these rules and obligations. Consequently, it isimportant for the doctors who are in charge of treating patients with biologicalmedicinal products to know the respective contributions of each available medi-cine, whether it is a reference product or biosimilar. This book is aiming atsimplifying the approach of products as complex as biosimilars. To do so, itreminds the reader:

• the complexity of biotechnology products and their mode of production;• factors of safety for their approval application and their marketing

authorisations;• the risk analysis and possibilities of interchangeability;• the French authorities’ interdiction of substitution of a prescription by the

pharmacist;• the analysis of the rules laid down by some learned societies or professional

associations, for a better follow-up and a better prescription of biosimilars.

Today, thanks to a centralised and specialised European licencing process, thenumber of biosimilars that got their marketing authorisation is limited butadvanced when compared to some countries, like the United States or Japan. Insome other countries, some companies offer copies of biotechnological productswithout having to apply for approval based on an approach equivalent to that ofEurope. Currently, Europe is essentially concerned by the growth hormone hae-matopoietic growth factor G-CSF (Granulocyte Colony Stimulating Factor) anderythropoietin. The next approval applications should be affected by the arrival ofcopies of other therapeutic proteins like insulin, interferon a and monoclonalantibodies. Similarly, copies of medicines other than therapeutic proteins shouldreach the biosimilars market, such as fragmented heparins; for which a recom-mendation by the licencing authority has been published. This sector of medicineis thriving. It requires a professional’s deep knowledge in order to maintain thequality and the safety of patient treatments.

Jean-Louis Prugnaud

Preface xiii

Contents

Biologicals’ Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1K. Ho and J.-H. Trouvin

From the Biosimilar Concept to the Marketing Authorisation. . . . . . . 23M. Pavlovic and J.-L. Prugnaud

Immunogenicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37J.-L. Prugnaud

Substitution and Interchangeability . . . . . . . . . . . . . . . . . . . . . . . . . . 47J.-L. Prugnaud

G-CSFs: Onco-Hematologist’s Point of View . . . . . . . . . . . . . . . . . . . 53D. Kamioner

The Oncologist’s Point of View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61C. Chouaïd

Biosimilars: Challenges Raised by Biosimilars: Who is Responsiblefor Cost and Risk Management? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71F. Megerlin

Afterword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

xv

http://dx.doi.org/10.1007/978-2-8178-0336-4_1

http://dx.doi.org/10.1007/978-2-8178-0336-4_2

http://dx.doi.org/10.1007/978-2-8178-0336-4_3

http://dx.doi.org/10.1007/978-2-8178-0336-4_4

http://dx.doi.org/10.1007/978-2-8178-0336-4_5

http://dx.doi.org/10.1007/978-2-8178-0336-4_6

http://dx.doi.org/10.1007/978-2-8178-0336-4_7

http://dx.doi.org/10.1007/978-2-8178-0336-4_7

Contributors

Christos Chouaïd Service de pneumologie Hôpital Saint-Antoine, UMR InsermS-707, 184 rue du Faubourg Saint-Antoine, 75571, Paris Cedex 12, France

Kowid Ho Département de l’évaluation des médicaments et produits biologiques,Agence Française de sécurité sanitaire des produits de santé, 143 BoulevardAnatole, 93200 Saint-Denis, France

Didier Kamioner Service de cancérologie et d’hématologie, Hôpital Privé del’Ouest Parisien, 14 avenue Castiglione del Lago, 78190 Trappes, France

Francis Megerlin MCF droit et économie de la santé Liraes, Université ParisDescartes, 4, avenue de l’Observatoire, 75006 Paris, France

Mira Pavlovic DEMESP Haute Autorité de santé, 2, Avenue du Stade de, Saint-Denis La Plaine Cedex, 93218 France

Jean-Louis Prugnaud Département de l’évaluation des médicaments et produitsbiologiques, Agence Française de sécurité sanitaire des produits de santé,Président Commission thérapie cellulaire et thérapie génique, 13 avenue JeanAicard, 75011 Paris, France

Jean-Hugues Trouvin Département des sciences pharmaceutiques et biologiques,Université Paris Descartes, Avenue de l’Observatoire 4, 75006 Paris, France

xvii

Biologicals’ Characteristics

K. Ho and J.-H. Trouvin

Introduction: From Generics to Biosimilars

In the field of chemical medicines (active substances derived from chemicalsynthesis), the ‘‘generic’’ procedure is well known and can be used once theprotection period has expired (patents and Marketing Authorisation). A ‘‘genericdrug’’ is a medicine with the same qualitative and quantitative composition inactive substance(s), the same pharmaceutical form as the reference drug, andwhose bioequivalence with the reference medicinal product has been demonstratedby appropriate studies. The Marketing Authorisation application file for a genericdrug is, compared to a new medicinal product, much more simplified insofar asthe usually required non clinical and clinical data are reduced to a comparativebioequivalence study versus the ‘‘reference’’ medicinal product.

Today, in the field of biological drugs, and more specifically of so-called‘‘biotechnological’’ drugs (see infra, definitions), patents and other data protectionare falling in the public domain (ex: Insulin, Somatropin, Erythropoietin, etc.).Mirroring the generic approach implemented for chemical medicines more thanthirty years ago, the question arises as to open the same possible development of‘‘copies’’ of these biological/biotechnological drugs and to have these copied drugsapproved, following the same simplified ‘‘generic’’ procedure.

K. HoDépartement de l’évaluation des médicaments et produits biologiques, Agence Française desécurité sanitaire des produits de santé, Boulevard Anatole 143, 93200, Saint-Denis, France

J.-H. Trouvin (&)Département des sciences pharmaceutiques et biologiques, Université Paris Descartes,Avenue de l’Observatoire 4, 75006, Paris, Francee-mail: [email protected]

J.-L. Prugnaud and J.-H. Trouvin (eds.), Biosimilars,DOI: 10.1007/978-2-8178-0336-4_1, � Springer-Verlag France 2013

1

For scientific and technical reasons that will be detailed hereafter, the simplifiedapproval application procedure—applicable to chemical generics—has not beendeemed fit for biological and biotechnological drugs and an alternative route hasbeen developed. The specific strategy, known as ‘‘biosimilar’’, for the develop-ment, assessment and licensing of those drugs claimed ‘‘similar’’ to a referencebiological product has been put forward to take into account these products’specificities and to ensure that the benefit/risk ratio of copies of reference bio-logical products has been correctly evaluated before they are put on the market.

Definitions

Biologics

A definition of a biological medicinal product is given in the European Directive2001/83/CE (Annex I). It meets two characteristics that justify the regulatoryrequirements and criteria applied to them when they are evaluated for the mar-keting authorisation application: ‘‘A biological medicinal product is a product thathas a biological substance as an active substance. A biological substance is asubstance produced or extracted from a biological source and that needs for itscharacterisation and the determination of its quality a combination of physico-chemical-biological testing, together with knowledge of the production processand its control.’’

Included in biological medicinal products are vaccines, drugs derived fromhuman blood and plasma, as well as any substance extracted from animal orhuman fluids or tissues, but also products known as ‘‘biotechnology’’ (see infra),and more recently, innovating therapy medicinal products; whose definition isgiven in the European Regulation 1394/2007.

The classification as a biological medicinal product leads to, for regulatorypurposes, the setting up of more demanding evaluation criteria; due to theseproducts’ complexity and their production processes that make their final qualitymore difficult to guarantee and master.

Genetically Engineered Products

Among active substances of biological origin is found a particular class ofproducts known as ‘‘genetically engineered’’, as detailed in European regulation2309/93, annex I, part A: Medicinal products developed by means of one of thefollowing biotechnological processes:• recombinant DNA technology;• controlled expression of genes coding for biologically active proteins in pro-

karyotes and eukaryotes, including transformed mammalian cells;• hybridoma and monoclonal antibody methods.

2 K. Ho and J.-H. Trouvin

In the context of biosimilars, genetically engineered products are mainly rep-resented by proteins known as ‘‘recombinant proteins’’, because they are expressedand produced by biological systems (bacteria, yeast, human or animal cells, insector plant cells, transgenic plants or animals) that have been genetically modified, byinserting specific genetic sequences coding for a therapeutically interesting protein(see infra, production process).

The most famous examples of ‘‘recombinant proteins’’ used in therapeutics areinsulin, growth hormone, erythropoietin, hematopoietic growth factors and, morerecently, monoclonal antibodies. It is on these recombinant proteins that the first‘‘biosimilars’’ approach has been designed.

Biological Products Complexity and Examples

The definitions of biological and biotechnological medicinal products, detailedabove, permit the identification of these products’ main characteristics thatunderlie the interrogations and technical difficulties routinely met to ensure aproduction as well as a consistent and reproducible quality control of theseproducts. These characteristics also explain why the simplified approach of genericdrugs is not strictly applicable to ‘‘copies’’ of biologicals.

A biological substance is inherently a complex molecular structure. Thiscomplex structure is difficult, even impossible, to get by chemical synthesis. So,whereas it is possible to chemically synthesize a twenty- or so amino-acid peptide,it is impossible to produce by chemical synthesis a protein as complex as coag-ulation factor VIII, growth hormone, or even insulin, though this one is a relativelysimple polypeptide.

This molecular complexity also explains the recourse to biological sources inorder to extract or produce biological substances. We’ll see below all structuralelements to consider for the characterisation, production, and quality control of abiological substance before qualifying it for therapeutic use and establishingacceptance criteria for each batch of medicinal product.

The second difficulty that springs from the complex structure of the molecule ofinterest resides in the technical analytical means to study in detail these variousstructural aspects of the molecule of interest (or of the molecular population),before considering an administration to the patient. As already stated, syntheticchemistry cannot produce these complex molecules. Consequently, in order toproduce these so-called complex molecules, ‘‘biological’’ production systems—with their stream of variability and complexity—will be put in place. Let’sremember that, in the definition of a biological medicinal product, the producingprocess is an integral part of the product (‘‘as well as knowledge of its manu-facturing process’’). In the field of biological medicinal products it is commonlysaid that ‘‘the process makes the product’’ or the process takes part in the productdefinition.

Biologicals’ Characteristics 3

We will successively consider the three elements of complexity, stated above, toillustrate what technical points have to be mastered before ensuring that a biologicalproduct, resulting from production system A, is identical (or declared identical) to asame biological product, this one resulting from production system B.

These complex elements will be studied while using the elements that conditionthe ‘‘quality’’ global profile of the final product in order to illustrate with preciseexamples a typical quality profile of a recombinant proteins. It is that ‘‘quality’’profile which also conditions the efficacy and tolerance profile of the medicinalproduct administered to patients.

Protein Molecular Complexity

In living matter, proteins are structures essential to the organisation of any organelleor organism; they play a central role in terms of structure (for instance, membraneproteins), of metabolic activity (enzyme, cytokine, hormone), or immunologic(immunoglobulins). This protein meets structural characteristics that condition itsbiological activity, the duration of its activity and presence in the organism (the termhalf-life is used), and at last its capacity to be recognized by the organism as a knownstructure (the self) or unknown (the non self) and trigger or not from the receivingorganism a defence reaction (neo-antigenicity risk that will be discussed later).

Concept of Primary, Secondary, Tertiary and Quaternary ProteinStructure

First of all, a protein defines itself by its primary structure (made of a chain ofamino acids) in a determined order (the protein sequence is under the sway ofdeterminism of the gene coding for that protein; mastering the gene coding for theprotein of interest will be mentioned in the production process).

The correct sequence of amino acids will determine the molecular constraints,with the consequence of forcing the protein to organise itself in a defined spatialstructure. The organisation is described in beta sheets and alpha helices that can berepeated and alternate all along the amino acids sequence (Fig. 1).

This spatial organization can be geopardized, depending on the physico-chemical surroundings to which the protein will be successively exposed, duringthe step of expression and production in the cell medium, and then during thepurification process (see infra, description of the production process) that may leadto pH or denaturing oxido-reduction conditions. Not to mention the various finalformulation steps (here we speak of the medicine preparation in its final form as aninjectable or freeze-dried preparation) that may lead to instability, cleavage, ordenaturing reactions.

Thus the intrinsic protein structure cannot be reduced to a sequence of aminoacids alone. Its spatial folding and the preservation of this folding must also betaken into account, as well as the spatial, possibly multi-meric organisation duringthe medicine’s all-life duration.


Figure 2 schematically summarises the four levels of structural organisation ofa protein molecule. This conservation of spatial structure, conforming to the‘‘natural mode’’ when it exists, is essential on one hand for maintaining the proteinbiological activity, and consequently for its therapeutic activity (for instance, apartially denatured enzyme will not develop its activity, as it is the case of acoagulation factor); and on the other hand, to ensure that the protein will not berecognized as ‘‘foreign’’ by the receiving organism. Indeed, a protein presentingone or several non-conformed parts of the molecule could induce a reaction ofdefence (immunogenic reaction) within the organism with formation of antibodiesdirected against the protein of interest.

Concept of Aggregates and Risks of Degradation

The sometimes complex spatial organisation of protein molecules exposes toanother risk of the protein non-conformity, that of aggregate formation.

The consequences of protein molecules aggregating between them are multipleand relatively difficult to predict and, sometimes, to detect. Their impact is,however, always negative for biological activity, as well as for clinical toleranceand immunogenic risk.

Fig. 1 Tri-dimensional rendering of a protein, with its conformational folding necessary for itsactivity. Let’s notice how complex this protein structure is, compared to the chemical molecule(in blue) on which the protein is active. Hemoglobin molecule on black background � PawelSzczesny # 7041692


Local environmental conditions (pH, ionic force, osmolarity, etc.) to be met bythe protein all along the production process, followed by purification, and finally ofpharmaceutical formulation are crucial. They will induce or not the aggregatingbehaviour or other degradation mechanisms. Thus, along the development of thisproduction/purification process, but also during the pharmaceutical formulation ofthe finished product (form under which the product will be ‘‘packaged’’ to beadministered), physico-chemical factors potentially inducing such a behaviour ofaggregating or instability will be studied in order to suggest systems and/orpharmaceutical formulations that limit these risks and ensure a stable quality fromone batch to the other, and stability in the period of claimed shelf-life.

Concept of Post-Translational Modifications: Glycosylation Example

A protein is not characterised only by its primary structure (correct sequence ofamino acids) or its spatial conformation (secondary to quaternary structure, Fig. 2);

Fig. 2 The four levels of protein organisation (figure adapted from Horton HR, Moran LA, et al.(2002) Principles of biochemistry, 3rd edition). Primary sequence, defined by the gene coding theprotein, gives its sequencing to amino acids building the ‘‘polypeptide chain.’’ Secondarystructure explains the polypeptide chain folding in space. Two types of structures may be noticed,one in alpha helix, the other in beta-sheet. Figure 1 shows an example of protein demonstratingsheets and helix alternating. The organisation in sheets and helix depends on the amino acidsequence (therefore on primary sequence), function of electrostatic forces and degrees of rigidityimposed by each of the amino acids. Tertiary and quaternary structures are adopted by proteinsfunction of their microenvironment and own chemistry. All proteins don’t adopt and don’tnecessarily need to be organised in quaternary structure. a Primary structure. b Secondarystructure. c Tertiary structure. d Quaternary structure


most often it has additional characteristics acquired during the cellular process ofprotein synthesis. These are called ‘‘post-translational modifications,’’ due to thefact that they occur once the gene (nucleic acids sequence) has been translated intothe corresponding protein sequence (the amino acid chain). These modifications arealso designated as ‘‘maturation phase’’ essential before the release/secretion of cellproteins. These modifications consist of the grafting on defined amino acids of oneor several chemical/biological groups; as, for instance, phosphate or sulphategroups, or sugars (then it is called glycosylation) that modify the global charge andphysico-chemical or biological characteristics of these ‘‘mature’’ proteins andcondition what they’ll become in the organism. It is important to remember thatthese post-translational modifications are occurring on specific sites of the protein(see below) are not controlled by the gene that expresses the protein sequence, theyare instead specific to each cellular kind (notably function of the enzymaticequipment of the cell line expressing the protein of interest and the cell cultureconditions (see infra, production process)). These sometimes complex chemicalreactions are thus not controllable by of mastering the gene sequence, but bymastering of production conditions in which the cell line chosen to express therecombinant protein will be put. Let’s notice that these maturation reactions areabsent inside prokaryotic organisms (bacteria), or very simple inside inferioreukaryotes such as yeasts. This notably explains that depending on glycosylationcharacteristics of the protein of interest, only a ‘‘mammalian’’ cellular system couldbe considered for production. To give an example, let’s detail the glycosylationreaction and its consequences upon characteristics and reproducibility of recom-binant proteins and clarify the possible consequences brought by a change of pro-duction system, (a change that may occur at all levels, from the nature of theproducing cell, to the conditions of culture, and finally of purification). Glycosyl-ation is the most frequent post-translational modification. The chemical modifica-tions introduced are very complex due to the glycanic structures that are added tothe protein skeleton. It is that complexity that adds to the global complexity andvariability of a mature protein. In a few words, let’s remind you what the proteinglycosylation step consists of in endoplasmic reticulum and Golgi apparatuses;specific structures of a cellular organisation. A glycosylation consists of branchingon the protein, on determined amino acids (for instance, for N-glycosylation, Asnwhich is in the Asn-X-Thr sequence), sugar groups such as mannose, fructose orgalactose following a well-determined orders (Fig. 3). These glycosylation chem-ical reactions will lead to the making of ‘‘sugar chains,’’ more or less complex anddiversified, considering all the possible attaching combinations (number of anten-na(e) on a glycosylation site, and the nature of sugars making up this antenna), evenif some mandatory sequences are found in each structure.

Finally, the end of the sugar chain is most often capped by a sialic acid inthe form of neuraminic N-acetyl acid (NANA) in Human cells, when for manymammals a part of the sialic acid is in the form of neuraminic N-glycolylacid (NGNA) because the gene which codes for the enzyme that allows theNANA form to become NGNA, is muted and inactive in humans. This speciesspecificity is to be taken into account when has to be chosen the cellular system of


expression/production of the recombinant protein of interest, to ensure that thesialylation is as close as possible to the human form. The mature protein, so‘‘glycolysed’’ and more or less ‘‘sialylated,’’ gets some characteristics that aremore or less acidic with a changed isoelectric point (pI). Consequently, at the endof post-translational modifications (maturation) and considering here the glyco-sylation step alone, the mature protein will show itself not as a single molecularform, but as a mix; a molecular population with the same basic protein structure(primary sequence imposed by gene sequence) on which various types of sugarchains will have been attached, giving then to each protein molecule its own pI.The analysis of this molecular population evidences not a single ‘‘form’’, but aseries of isoforms on which a qualitative and quantitative study can be performedby appropriate analytical techniques that separate the various isoforms; forinstance, according to their charge.

Generating different isoforms, of which relative proportions are functions of thecell culture medium conditions, induces the notion of a protein ‘‘microheterogeneity.’’So, thanks to these ‘‘post-translational modifications’’ specific to the protein of interest(but also to the expression/production system, as well as the purification scheme),a protein is characteristic because of its ‘‘glycosylation profile;’’ described most oftenby a series of visible and quantifiable bandwidths, by separation methods ofisoelectrofocalisation type. Preserving the glycosylation profile will guarantee thebiological activity and the tolerance profile in patients.

Post-translational modifications, usually illustrated by the glycosylation profile,are intrinsic quality criteria of the protein as well as critical parameters to consider

sialic acid

gal

glcNAc

man

fuc

glycans

protein

carbohydrates

Fig. 3 Schematic drawing of carbohydrate residues (or glycanic structures) present on someprotein sequences. Glycanic structures are obtained by combining the sugar group’s natureGal = Galactose, Man = Mannose, Fuc = Fucose, glcNAc = N-acetyl glucosamine and itsorganisation in antennae (mono, bi, even tri-antennae). Let’s also note the presence of a ‘‘sialicacid’’ group that sometimes caps antennae’s ends. The sialic acid groups are notably contributingto the protein molecule’s half-life


during the assessment of the production process and its reproducibility, notablywhen changes are introduced in the production method, and a fortiori when a newmanufacturer offers a ‘‘biosimilar’’ version of a reference protein.

Indeed, for the new producer of a given glycosylated protein, one could fear anisoform distribution different from that of the original molecule. This differentisoelectric profile—sometimes hard to distinguish by the only analytical methodsoffered by the manufacturer, will potentially have an impact on the pharmacoki-netics or the biological activity of the therapeutic protein. Then it will be thepharmacological and/or clinical data that will reveal the sometimes subtle changein isoform distribution when the quality control analytical data are detecting nonoticeable difference.

Although some studies suggest that the consequence of a different isoelectricprofile mostly concerns the neo-antigenicity risk, it seems that this phenomenonrather impacts the half-life of the molecule which will be more or less rapidlyeliminated by the receiving patient body. Indeed, the sugar chains, notablydepending on their sialic acid capping, protect the protein from capture anddegradation by hepatic cells.

Thus a recombinant protein will have to have an adapted glycosylation, as wellas a correct sialic acid level (in the NANA form), to not to be eliminated tooquickly and keep a sufficient pharmacological activity and reduce any potential togenerate in patients a defence reaction with formation of antibodies to the proteinof interest.

Other Modifications Linked to the Process and/or Conservation/Formulation

In the paragraphs above, we have mentioned the different critical parameters of amedicinal protein molecular structure. These functional attributes (primary andspatial structures, state of non-aggregation, glycosylation profile and other post-translational chemical modifications) are to be considered along the whole chain ofproduction and then of conservation of the biological medicinal product. Theseparameters must be monitored during the final production step of the pharmaceuticalform that follows the active substance’s production/purification. That step of pro-ducing the pharmaceutical form (finished product to be administered to the patient)must also respect the molecular structure, without aggression or degradation.

Then, at this stage of the finished product, the same kind of difficulties arisewhen a medicinal product known as ‘‘biosimilar’’ and its reference product havedifferent pharmaceutical formulations and/or different presentations.

Any change in the excipient formula may indeed modify the stability profile ofthe molecule, which will be more rapidly degraded, or will give birth to impuritiesor degradation products, notably, aggregates.


Molecular Complexity

In conclusion of this chapter dedicated to protein molecular characteristics, oneshould above all remember that biological products are complex structures, not onlybecause of their basic protein structure, but also because of other modifications thatthey undergo during their maturation, generating a ‘‘final form’’ that is not a ‘‘sin-gle’’ and monomolecular entity (as could be expected of a chemical molecule with99.9 % purity) but rather a complex mix of the same protein molecule under variousstructurally close isoforms. Here one speaks of intrinsic ‘‘microheterogeneity’’ of abiological substance. The microheterogeneity will be like a digital print of thetherapeutic protein which will also be predictive of the protein’s activity andtolerance profile when administered to the patient. Products with a close structurethat have not been eliminated during the purification steps must be added to this mixof isoforms (when the protein presents a glycosylation profile), as well as impuritiesbrought by the various reagents and different steps of the purification process and theproducing of pharmaceutical form (see infra).

This mix will ultimately be considered as the product of interest that will bequalified in its global form by clinical data. It has to be considered that, once this‘‘mix’’ is characterised and validated by clinical use, the producer will have tomake every effort to do ensure that, batch after batch, the product delivered to thepatient meets the same quality criteria.

Figure 4 attempts to summarize the different sources of variability and thekeeping of elementary characteristics needed for the biological activity of acomplex structure of a biological substance, such as a protein.

Variability

Structure III & IV

Post - translational modifications Sequence

AA composition and sequence Source of variability

Substitution Oxidation Deamidation Truncated form N - and C - terminal heterogeneousness

…

Type of modification Glycosylation ( – N & O - linked) Methylation / Acetylation / Acylation Phosphorylation / Sulfatation

…

Type of variability Conformation Aggregates dissociation

…

Fig. 4 Molecular structure: a source of variability and heterogeneity


The Analytical Challenge

As explained above, a biological substance has a complex chemical structure andis rarely (if ever) a pure and single molecular entity, but rather a ‘‘molecularpopulation’’ that includes (see Fig. 5):• the majority molecular form and its variants and isoforms, each carrying an

intrinsic biological activity close to the majority molecular form’s biologicalactivity (for instance EPO and its isoforms);

• the impurities linked to the product itself, but which practically carry no bio-logical activity;

• the impurities linked to the production process or to the purification scheme.It is this mix of chemically complex structures that has to be controlled,

qualitatively and quantitatively, before considering the release of the medicinalproduct for its clinical use.

These controls are aimed at verifying, compared to preset specifications, thatthe medicinal product is the one expected for the patient, at the right dosage, andthat the level of impurities conforms to what is expected.

One then understands that the physico-chemical and biological analysis of thatcomplex mix, (to evaluate different distinct molecular properties that are comple-mentary and necessary) cannot be done by a single method, and that, as in the definitionof a biological product, ‘‘a combination of physico-chemical and biological tests’’ willbe needed to be in a position to extensively characterise the protein of interest in termsof mass, charge, spatial folding and multicatenary or multimeric and finally its post-translational characteristics (quantitative analysis of different isoforms).

There are numerous analytical methods to evaluate the purity and impurityprofile of biological medicinal products; among which there are circular dichroism,nuclear magnetic resonance, immunological tests (ELISA, immunoprecipitation,biosensors, etc.), biological activity on in vitro models (cell culture) and in vivo(animal models), various chromatography techniques (HPLC peptide mapping),

Peptid variants

Protein of interest

Post-translational variants

PURITY PROFILE

IMPURITY PROFILE

degradationProcess-related impurities

3D structure

Fig. 5 Schematic representation of the notion of purity/impurity profile


electrophoresis methods (SDS-PAGE; IEF; CZE), static and dynamic light diffu-sion mass spectrometry, X-ray techniques, etc.

These methods are in constant evolution in their sensitivity levels, discrimi-nating capacities and specificity. They allow for a continuous improvement ofknowledge and understanding of these products. However, these tools reach theirlimits to exhaustively report a biological product quality profile. This fact has beenillustrated several times for biological medicinal products, at different stages oftheir development (before or after approval), for which variants or impurities havebeen detected ‘‘at a late date.’’ In some cases, these variants or impurities that havealways been present in the product, are incidentally discovered because ofimproved analytical methods and have no clinical consequences. In other instan-ces, impurities have been looked for and detected after pharmacovigilance signals,such as undesirable effects observed in the course of a clinical trial or a long termtreatment by these products.

The methods are many and each shows a more or less fine analysis of the variouscharacteristic of the product of interest. The methods may be classified in largefamilies, according to molecular characteristics they are capable to study (Table 1).

Today if many molecular characteristics can easily be investigated and verified(molar mass, glycosylation profile), there are still three-dimensional structuralelements that are especially telling criteria, but difficult to analyse in case of complexproteins. The current physico-chemical methods (circular dichroism, near andfar UV spectrometry, NMR, etc.) can compare only certain aspects of the proteinthree-dimensional structure to the given reference protein. Consequently, in the

Table 1 Main analytical methods that can be used in the study and control of proteins of othermacromolecules

Method Size Charge Primary structure 2�/3�Structure Purity Potency

HPLC

Exclusion size +++ – – ++ ++ –

Ions exchange – ++++ +++ – – –

Reverse phase +++ +/– +++ ++ ++ –

Electrophoresis

SDS-Page +++ – +++ – +++ –

IEF – +++ + ++ +++ +

W-Blot +++ – ++ +++ +++ –

Dosages

Immuno-assays – – +/– +/– – ++

Receptors fixation – – ++ +++ – ++

In vivo dosage – – +++ ++++ +/ +++++

Each method covers all or part of the analysed molecules structural parameters. Data provided bythe various analytical methods (adapted from Thorpe R, personal communication)


structural analysis of the proteins, some characteristics necessary for the proteinbiological activity will not be verified by the only physico-chemical methods, and itis routine to perform, on top of physico-chemical tests, so-called biological tests(that involve reactions of the antigen–antibody type, or agonist-receptor, or recog-nition of an enzymatic site, even tests for pharmacological activity on animals) toverify the molecular global integrity of the molecule.

This analytical challenge also explains that it is sometimes difficult to claim theabsence of difference when two molecules are compared. This difficulty to draw aconclusion on a difference or absence of difference explains itself either by inade-quate methods or a detection limit reached. In a demonstrated absence of différence,there is a risk to wrongly draw the conclusion that the two compared products are‘‘identical’’. In this field of comparing molecules, it is important to remember thatthe absence of proof is not proof of absence; there have been described secondaryeffects in patients who had received molecules derived from a modified productionprocess, and these molecules had been successfully analysed in search of structuralor molecular modifications that the changed process could have induced. It has onlybeen shown after administration to patients that the molecule, produced by amodified process, presented a different behaviour and notably led to the productionof antibodies (immunogenicity profile) against the protein of interest, whereas all theanalytical parameters concluded for an equal (identical) quality and purity profile.

The Production Process Challenge

An active substance known as ‘‘biological’’ is by nature complex; consequently itsproduction has to make use of processes that are themselves complex, using livingmaterials and reagents with steps of fermentation/cell culture followed byextraction/purification from a complex matrix composed of the chosen expressionsystem metabolism. We shall describe the production process for a recombinantprotein, in order to illustrate the ‘‘critical’’ points that will condition the finalquality of the protein produced.

It should be briefly recalled that in a process of production called ‘‘biotech-nological,’’ it is a cell system that ensures the expression of a protein of therapeuticinterest. This cell system (bacteria, yeast, insect, plant or mammalian cell)undergoes a genetic modification which consists in the insertion of a foreign geneinto the genome of the host cell—one speaks of transgene—coding for the proteinof interest. The protein so produced (following the genetic sequence that will havebeen inserted into the host cell system and after post-translational maturation thatthe host system is capable of making) will have then to be collected in this culturemedium to be purified to the point of getting a protein solution at a degree of purityclose to 100 %. It is that purified protein that will be transformed (formulated) intothe final pharmaceutical form (most often injectable).

So, to produce a recombinant protein, and get a biotechnological medicinalproduct, several steps must be identified; each step has an impact on the robustnessof the production process, and ultimately upon the quality of the protein of interest:


• developing the cell system expressing the transgene that will have been insertedinto its genome;

• implementing the culture of the genetically modified cell;• ‘‘collecting’’/harvesting from the production system;• purification of the protein by different chemical or biological steps in order to

get an active substance with a declared and validated purity level;• pharmaceutical formulation to get the medicinal product in its final form.

We will briefly introduce each of these steps, while putting the accent on thecritical point of each of them, since these critical points could fail at any momentand lead to the production of a lesser quality protein or of a protein different fromthe expected one. In the context of biosimilars, we will see that the process,implemented by a second manufacturer, is therefore one of the essential points tomaster the quality of a ‘‘copy’’ product.

Developing an Expression System: From Genetic Codeto Medicinal Protein

Genes are DNA portions carrying a message that ultimately leads to the production ofproteins. They are present in all living creatures’ genomes and are sequences ofnucleotides (A, T, G and C). Each of these genes’ sequence is specific of a protein.The cells’ machinery transcribes the genes (DNA) into RNAm which in turn aretranslated into proteins. These four steps are represented in the following sequence:

DNA ? transcription ? RNAm ? translation ? native protein ? post-trans-lational modification ? mature protein ? excretion/secretion ? release of theprotein into the extracellular medium.

Pre-Production Step, Development of Expression System,Concept of Genetic Engineering

The gene encoding the protein sequence of interest is the first key element of thissystem since only a correct gene sequence will give rise to the expression of aprotein with the right conforming primary sequence.

The gene, most often of human sequence, must be inserted into the cell that willhave been chosen as the production host system. As the genetic code is universal,it will be read the same way by all cellular systems of the animal, plant or bacterialkingdom (even as the existence of dominant codons per cell system is known).

This universality is the basis of the production of recombinant1 therapeuticproteins of human sequence into heterologous host systems (bacteria, yeast, plant,mammalian cell, transgenic animals) to make that host system ‘‘produces’’ aprotein of given sequence.

1 ‘‘Recombinant’’ means a foreign DNA integration into the genome of a host that becomesgenetically modified.


Biotechnology is therefore aimed at inserting a transgene encoding the proteinto be produced in a ‘‘host’’ producer organism named expression system.

This latter is organised in cell banks (master and working cell banks, thatsystem will be described later, Fig. 6) to ensure the production system repro-ducibility. Cells issued from cell banks are cultured to produce the protein ofinterest. Then, this protein is extracted from the culture medium, purified, andformulated into a finished product which is the marketed product.

Concept of Expression Cassette and Vector

For the human protein to be correctly expressed by the producer host organism, it isnecessary to optimise the gene sequence in using for each amino acid the dominantcodon of the species used. The optimised sequence is affixed to a promoter whichcontrols the protein expression by the genetically modified host system. Selectingthe promoter depends on the host cell and leads to optimising the expression yield.

At that early stage of ‘‘genetic construct,’’ selecting the producer system con-ditions the nature of the protein produced and related substances which constitute‘‘the active substance of interest.’’

In other terms, a genetic construct, developed for a given system of production,constitutes a first element of originality and of characterisation of the producedprotein.

Expression plasmid Host cell

Selected clone of thegenetically modified

Host cell

Cell expansion; quality controls and repartitio incryotubes

The cell bank system

First level of the « cell bank system »The Master cell bank (MCB)

consisting of a hundred of cryotubescontaining the Clone of interest

Each cryotube contains ca. 10E6genetically modified host cell

Amplification of one cryotube, in defined cell culture conditionsgives rise to a new cell population to be distributed in cryotubes

which constitute the « working cell bank »

Second level of the « cell bank system »The Working cell bank (WCB)

A production run is initiated by expandingone cryotube of the WCB

Harvest and then dowstream process

Some cells are sampled at the end of the cell culture.These end of production cells (EOPC) are then further

amplified to give rise to the late expanded cell bank

Late expanded cell bank (LECB)

Fig. 6 The two-tier system of cell bank


Any other system of production, developed by a ‘‘biosimilar’’ manufacturer forinstance, is by nature different in its design and its construction and therefore in themost intrinsic molecular characteristics of the biological substance produced.

Host Cell and Expression System

Selecting the host cell and the expression system is above all conditioned by thenature of the protein to be produced. In fact, depending on the protein structurecomplexity and the necessity (or not) of a post-translational modification step, thedifferent cell systems may or may not meet the requirements. Typically there arefive types of organisms that may be used in the production of recombinant proteinsof therapeutic interest: bacteria, yeast (and fungi), mammalian cells (human cellsincluded), insect and plant cells. To these five main expression systems in vitro,may be added two in vivo expression systems, with transgenic plants and animals,even if the use of these systems is still anecdotic.

It is important to remember over all that, for a given system, the productobtained may notably differ from the others because of its many impurities and thedistribution of isoforms and other structural variants that have been stated above.We will briefly discuss some aspects specific to these different expression systems.

Bacteria have been the first in vitro system of production of recombinant proteins,notably Escherichia coli. This system combines an easy use and good yields;however, its use is limited to the production of proteins that don’t necessitate a post-translational modification, since prokaryotic cells are not equipped in enzymesneeded for glycosylation. During the protein production phase, the bacteria willeither produce it in ‘‘inclusion bodies’’ inside the cell, or by internal secretion in theperiplasmic space. The formation of inclusion bodies produces protein semi-purifiedfractions that are easily collected by a simple extraction. However, this extractiontechnique supposes the proteins’ resolubilisation in a solvent such as urea, whichinevitably implies their partial or total denaturing. This denaturing must then befollowed by a renaturing phase of the protein in order for the protein’s spatialstructure to be regenerated to ensure its biological activity (see supra). This phase ofdenaturation-renaturation is tricky and if it is not well performed or mastered, maylead to the formation of a more or less consequent fraction of protein molecules withabnormal conformations, which could be responsible for reactions from the recipientorganism which don’t recognize the denatured protein structures (neo-antigenicity).

Eukaryotic Systems

For complex proteins, necessitating post-translation specific reactions (such asspecific conformations, oligomerisation, proteolytic cleavages, phosphorylationsor glycosylation reactions) have to be considered production systems able toprovide a machinery needed for these maturation reactions, which generally takeplace in the endoplasmic reticulum and in the Golgi apparatus of eukaryotic cells.


Three levels may be identified within eukaryotic systems:• lower eukaryotes with yeasts and fungi. These systems are able to make rela-

tively simple post-translational modifications, while preserving their simpleculture conditions and a production yield close to bacterial systems’. However,this system remains limited to simple glycoproteins since yeasts produceN-glycosylations rich in mannose residues; strongly immunogenic for humans;

• higher eukaryotes with mammalian cells, but also insect or plant cells. In gen-eral these cells are used in industrial production only when microbial or fungalare deemed not adapted to produce the protein of interest. In fact, unlike thesystems above, higher eukaryotes are systems more difficult to industriallyimplement and with a lower yield (as compare to bacteria or yeast), and veryconstraining culture conditions. Mammalian cells such as ovarian cells ofChinese hamsters (Chinese Hamster Ovary (CHO)) are commonly used toproduce complex glycoproteins. More recently, human cell lines have beendeveloped and qualified (osteosarcoma 1080 or a line of human embryo kidneyHEK293 (Human Embryonic Kidney)). These human cell lines have the mainadvantage of being able to make post-translational modifications of humanstructures, which could be interesting, notably when the glycosylation profilemust be close to the ‘‘natural’’ profile (for instance important for the pharma-cokinetic profile);

• at last, a third expression/production system has been recently developed, withplant and animal transgenesis: to have produced by a genetically modified plantor animal, the therapeutic protein, in a tissue (for the plant) or in a fluid (mostoften milk for transgenic animals) easily accessible for the ‘‘harvesting’’. Fromthis harvest, protein purification and final product formulation follow the pro-duction diagram that will be described below.

Concept of Cell Bank

To ensure homogeneity and reproducibility in the protein of interest production, asystem of ‘‘cell banks’’ (also designated as two-tiered seed lots systems (Fig. 6)) isset up starting from the initially selected producing clone. For that, the clone ofinterest (obtained after genetic modification of the host cell line) is put into culture,and the produced population is aliquoted in fractions (a few million cells percryotube), then cryo-conserved. This first batch of tubes (generally around onehundred of cryo-tubes) constitutes the Master Cell Bank (MCB) (since it is directlya product of the selected clone amplified). To ensure the sustainability of thisMCB, a Working Cell Bank is prepared, by amplifying one or two MCB cryotubeswith distribution of cells produced in a working cell bank (WCB) consisting alsoof hundreds of cryo-tubes, as with the MCB. Each production batch is initiatedfrom a WCB cryotube. It is when, after several production batches have beenmade, the WCB decreases and a new WCB may be prepared, according to the


same process as described above, by taking, in the MCB, one or two cryo-tubes.This ‘‘two level’’ cell bank ensures the sustainability of the cell line that has beenbuilt from the same genetically modified cell.

Active Substance Production: Fermentation or Cell CultureConditions

Once the expression system built up, i.e. the host cell genetically modified and theproducer clone set up in a bank, it is time to proceed with the culture of the host cell inorder that it produces, in specific and optimised culture conditions, the protein of interest.

These organisms’ culture is carried out in fermenters (bacteria and yeasts) orcell culture systems (roller bottle, cytocultivator, hollow fibres, etc.).

It would take too long to detail the different technical solutions that have beendeveloped in order for cells maintained in artificial conditions of production to, notonly survive, but above all produce in notable quantities the protein for which theyhave been genetically modified.

Culture conditions, i.e. the way the host cell is maintained, fed, oxygenated, etc.are determining elements for production yields and the intrinsic quality of theproduced protein. To show how important and critical the culture conditions are toobtain a determined quality protein, let’s remember that quality and intensity of theglycosylation made by the eukaryotic cells are notably dependent on the cultureconditions to which the host cell is exposed. Therefore one may observe that a pHchange, a modification of oxygen partial pressure, or a change in the speed of carbonhydrates supply may lead to a modified distribution of glycosylated isoforms.

Thus, for a protein of interest, the production process is a critical step to‘‘harvest’’ ultimately a bulk product that has to be as reproducible as possible,notably in terms of production impurities, products derived of cell metabolism, andobviously isoforms or molecular variants.

In the context of the production of a protein called ‘‘biosimilar’’, it is under-standable why the development of a production process as close as possible to theprocess used for the reference protein production, and the mastering of this pro-cess’ reproducibility are key elements to ensure that the biosimilar protein will be,in fact, as similar as possible to the reference protein, in all its structural andphysico-chemical aspects. These same questions are evidently implied if theconsidered production process makes use of transgenic animal or plant, with addedelements of complexity and qualification, that we cannot describe here.

Purification

Once the expression and production phase done (whatever the production system,either in vitro or in vivo), it is time for ‘‘harvesting’’ the medium in which lies theprotein (bacteria’s all body, culture juice of yeast or cells, transgenic animal’s


milk, etc.) and then to proceed with the steps of extraction and purification of theprotein of interest.

The objective of this so-called ‘‘downstream process’’ is to extract from thiscomplex biological matrix the protein of interest and eliminate all the inherent‘‘contaminant’’ substances, notably those coming from the production cell (DNAand proteins), the used raw materials (reagents, culture media…), and fromdegradation products. A second purification objective is to eliminate, throughdedicated steps, pathogens potentially transmittable and brought in by the different‘‘biological’’ elements introduced all along the process. Worth mentioning here isthe risk of contamination by virus from the cellular system, or other adventitiousagents such as the agent responsible for spongiform encephalopathies, etc.…

There are numerous strategies to extract and purify proteins, each one pre-senting a certain level of specificity (in order to select only the protein of interest),of yield (amount of proteins of interest eliminated with effluents), and of main-taining the molecular integrity of the protein being purified.

The purification system may also introduce differences in the protein qualityprofile, between production batches or between producers in:• qualitatively and quantitatively selecting isoforms;• co-purifying different impurities;• triggering, thanks to more or less drastic purification conditions (notably sol-

vent/detergent treatments typically applied for viral elimination/inactivation),denaturations/degradations of the purified molecules.At the end of the downstream process, there is a said ‘‘purified’’ protein with a

level of purity that must be qualified and verified in comparison with preset criteriaor specifications (see supra, analytical challenge and quality control strategy).

At this stage of the process, the protein is considered as ‘‘the active substance’’that can be stored for conservation before producing a pharmaceutical form. Infact, the protein has not yet a ‘‘medicinal product’’ form, meaning a form thatcould be administered to the patient. It is then time to proceed with the ‘‘bulkprotein’’ towards the last step of ‘‘making a pharmaceutical form’’, called theformulation of the finished product.

Towards a Pharmaceutical Form

Making a pharmaceutical form consists of inserting the protein of therapeuticinterest in a medicinal form that will be administered to the patient.

As they are proteins, these substances cannot (except for rare exceptions) beadministered orally. Therefore the medicinal product to prepare will be adminis-tered by injection (subcutaneous, intramuscular or intravenous).

For that, a formulation must be done with excipients able to ensure the proteinbest possible for dissolving, and maintaining its physical integrity (for instance noaggregate forming) and chemical integrity (no alteration of chemical propertiessuch as oxidation, reduction, loss of groups of glycosylation, etc.), since all these


degradation reactions are mainly controlled by chemical conditions (pH, ionicstrength, humidity, etc.) and/or physical conditions (storage temperature, liquid orsolid state) imposed upon the protein. This step of formulation and excipientselection and of conditions of making a pharmaceutical form allows to ensure thecompatibility of the active substance with its final administration environment, butalso to guaranty the active ingredient stability inside the medicinal product duringits manufacturing process as well as its storage all along the shelf-life.

Thus, this last step must be integrated in the whole process of the medicine’sproduction as a possible source of technical obstacles to achieve a medicinalproduct of consistent quality.

Among examples in which the medicine’s formulation had a consequence interms of serious undesirable effects, let’s mention the case of severe anemia (purered cell aplasia) due to the formation of antibodies anti-active substance, reportedafter a change in the formulation of an erythropoietin that was on the market,without noticeable pharmacovigilance signal, for 10 years.

This example, like many others, shows that maintaining the molecular integrityof the protein of interest is resulting from the full mastering of all the process’steps, from the host cell system construction, to the culture of that cell system, thenextraction/purification, and finally the making of a pharmaceutical form andcompliance with the medicinal product handling and storage conditions.

Conclusion

All along this chapter, we have listed and presented the different scientific andtechnical elements linked to the very nature of biological substances, their com-plex molecular structure, their production and purification process, their makinginto a pharmaceutical form, explaining why there are so many unknowns andso many sources of difficulties in making an exact ‘‘copy’’ of the molecule ofreference. It can thus be concluded that the ‘‘generics’’ approach is not applicableto scientifically and sufficiently guaranty the ‘‘copy’’ biological will be of the samequality, safety and efficacy profile as the reference medicinal product againstwhich the ‘‘copy’’ is claimed to be similar.

Only a careful comparison of the two products (the reference and the copy),of their production conditions, of control and storage- reinforced by non clinicaland clinical results—will give a chance to the health authorities to guarantee that,within the limits of scientific knowledge, and within the limits of the investigationsthat had been conducted, there is no possible evidencing of significant differencesor potentially generating different efficacy and safety profiles, and that on thisbasis, the two products are considered as ‘‘biosimilars’’.


Further Reading

• Directive 2004/27/CE du Parlement européen et du conseil (31 mars 2004)• EMEA/CHMP/31329/05 Annex Guideline on Similar Biological Medicinal

Products containing Biotechnology-Derived Proteins as Active Substance:Non-Clinical and Clinical Issues. Guidance on Biosimilar Medicinal Productscontaining Recombinant Granulocyte-Colony Stimulating Factor

• EMEA/CHMP/31329/05 Annex Guideline on Similar Biological MedicinalProducts containing Biotechnology-Derived Proteins as Active Substance:Non-Clinical and Clinical Issues. Guidance on Biosimilar Medicinal Productscontaining Recombinant Granulocyte-Colony Stimulating Factor (CHMPadopted February 2006)

• EMEA/CHMP/32775/05 Annex Guideline on Similar Biological MedicinalProducts containing Biotechnology-Derived Proteins as Active Substance:Non-Clinical and Clinical Issues. Guidance on Similar Medicinal Productscontaining Recombinant Human Insulin

• EMEA/CHMP/32775/05 Annex Guideline on Similar Biological MedicinalProducts containing Biotechnology-Derived Proteins as Active Substance:Non-Clinical and Clinical Issues. Guidance on Similar Medicinal Productscontaining Recombinant Human Insulin (CHMP adopted February 2006)

• EMEA/CHMP/42832/05 Guideline on Similar Biological Medicinal Productscontaining Biotechnology-Derived Proteins as Active Substance: Non-Clinicaland Clinical Issues

• EMEA/CHMP/42832/05 Guideline on Similar Biological Medicinal Productscontaining Biotechnology-Derived Proteins as Active Substance: Non-Clinicaland Clinical Issues (CHMP adopted February 2006)

• EMEA/CHMP/437/04 Guideline on Similar Biological Medicinal Products• EMEA/CHMP/437/04 Guideline on Similar Biological Medicinal Products

(CHMP adopted September 2005)• EMEA/CHMP/94526/05 Annex Guideline on Similar Biological Medicinal

Products containing Biotechnology-Derived Proteins as Active Substance:Non-Clinical and Clinical Issues. Guidance on Similar Medicinal Productscontaining Recombinant Erythropoietins

• EMEA/CHMP/94526/05 Annex Guideline on Similar Biological MedicinalProducts containing Biotechnology-Derived Proteins as Active Substance:Non-Clinical and Clinical Issues. Guidance on Similar Medicinal Productscontaining Recombinant Erythropoietins (CHMP adopted March 2006)

• EMEA/CHMP/94528/05 Annex Guideline on Similar Biological MedicinalProducts containing Biotechnology-Derived Proteins as Active Substance:Non-Clinical and Clinical Issues. Guidance on Similar Medicinal Productscontaining Somatropin

• EMEA/CHMP/94528/05 Annex Guideline on Similar Biological MedicinalProducts containing Biotechnology-Derived Proteins as Active Substance:


Non-Clinical and Clinical Issues. Guidance on Similar Medicinal Productscontaining Somatropin (CHMP adopted February 2006)

• EMEA/CHMP/BWP/49348/05 Guideline on Similar Biological Medicinal Prod-ucts containing Biotechnology-Derived Proteins as Active Substance: Quality Issues

• EMEA/CHMP/BWP/49348/05 Guideline on Similar Biological MedicinalProducts containing Biotechnology-Derived Proteins as Active Substance:Quality Issues (CHMP adopted February 2006)

• EMEA/CPMP/3097/02 Note for Guidance on Comparability of MedicinalProducts containing Biotechnology-derived Proteins as Drug Substance: NonClinical and Clinical Issues (CPMP adopted December 2003)

• EMEA/CPMP/BWP/3207/00 Rev.1 Guideline on Comparability of MedicinalProducts containing Biotechnology-derived Proteins as Active Substance-Quality Issues (CPMP adopted December 2003)

• Thorpe R et al (2005) Clinical and Diagnostic Laboratory Immunology 12:28–39


From the Biosimilar Conceptto the Marketing Authorisation

M. Pavlovic and J.-L. Prugnaud

Introduction

The concept of similar biological medicinal products similar to a referencebiological medicinal product has been recently introduced in the European legis-lative framework. As it has been stated in this book’s introduction, the biosimilarterm designates in common language the ‘‘copy’’ concept of a biological medicinalproduct. The purpose was to open a regulatory route for pharmaceutical companieswilling to develop biosimilar medicinal products once the marketing protection ofthe ‘‘reference’’ biological medicinal product expired.

Several medicines of this particular field are or will have expired patents in thenear future, which offers pharmaceutical companies the possibility to developsimilar products and to obtain the same therapeutic indications as the referenceproducts.

Even if this strategy could be easily assimilated to the standard generic approach(which, for a chemically derived substance, requires a single demonstration of bio-equivalence with the reference product), the generic approach was not consideredadequate to establish the quality, safety, and efficacy of biosimilars. That is due to the

Pavlovic article M, Girardin E, Kapetanovic L, Ho H et Trouvin JH, Similar Biological MedicinalProducts Containing Recombinant Human Growth Hormone: European Regulation. Horm Res2008;69:14–21. � 2007. Karger AG, Basel.

M. Pavlovic (&)DEMESP Haute Autorité de santé, avenue du Stade de 2,93218 Paris, Francee-mail: [email protected]

J.-L. PrugnaudAgence Française de sécurité sanitaire des produits de santé, Département de lévaluation desmédicaments et produits biologiques, Président Commission thérapie cellulaire et thérapiegénique, 13 avenue Jean Aicard, 75011, Paris, France


23

complexity of the biotechnology-derived products themselves as well as their man-ufacturing process. In most cases, molecular complexity and heterogeneity inherent tobiological products do not allow for their full and guarantied characterization.

The quality attributes of the active ingredient are highly dependent on itsmanufacturing process and any change in the manufacturing process may affect thequality attributes and impact on the safety or efficacy profiles of the product.Therefore the European legislation has provided a specific regulatory framework(called ‘‘biosimilar approach’’) for biological medicinal products similar toreference biological medicinal products. It is applicable to any biological medic-inal product, which confers an originality to the European regulation and its uniquecharacter. Practically, the biosimilar approach developed in the recommendationsfor approval application apply to well-characterized recombinant proteins, such asinsulin, somatropin, erythropoietin, G-CSF (Granulocyte Colony StimulatingFactor). Other recommendations have been issued for low molecular weightheparins and alpha interferon, or are being drafted for monoclonal antibodies.These guidelines are made by the CHMP (Committee for Medicinal Products forHuman Use) of the European Medicines Agency (EMA); they are relevant toquality, non-clinical and clinical issues to be developed in order to be addressedfor a biosimilar approval application.

Definition of Biosimilars

‘‘When a biological medicinal product similar to a reference medicinal product doesnot meet the conditions stated in the generics definition, notably because of differ-ences linked to raw material or differences between manufacturing processes of theproduct and the reference product, appropriate preclinical or clinical studies relatedto these conditions must be provided.’’ [European guideline 2004/27 art.10 (4)].

In this chapter the general recommendations will be summarized and analysedin relation to the development quality of a biosimilar, followed by those related topreclinical tests necessary before the first human administration. The generalrecommendations for a clinical evidencing of similarity in terms of safety andefficacy will be particularly developed for biosimilars used in oncology andhaematology. These essentially concern erythropoietin and the growth factor G-CSfor which the first biosimilars have been put on the market.

Pharmaceutical Authorisation Background

General recommendations appear in a general text on biosimilars, which introducesthe concept of biosimilarity and gives a definition of the main principles of bio-similars development in terms of quality, safety and efficacy.1 A company

1 Guideline on Similar Biological Medicinal Products. CHMP/437/04 (CHMP adoptedSeptember 2005).

24 M. Pavlovic and J.-L. Prugnaud

developing a medicinal product similar to a reference biological medicinal productmust choose the reference among medicines authorized by a complete file with theEuropean Community. The concept of a biosimilar is applicable to any biologicalmedicinal product. However, in practice, demonstrating the similarity will dependon a possible complete characterisation of the product. For that it is necessary tohave not only data on physico-chemical and biological properties, but also to knowthe manufacturing process and its controls. As minor changes in this manufacturingprocess may alter the product at a molecular level, the biological product safety andefficacy profile depends on the robustness and follow-up of quality issues.

The biosimilar approach takes into account the following points:• the ‘‘standard’’ generic approach is not considered as acceptable. The biosimilar

approach is based on exercises of comparability due to the complexity ofbiotechnology-derived products;

• exercises of comparability can only apply to highly purified products that maybe correctly characterized. It is not always the case, notably for extractionproducts with biological sources, or those for which only a limited clinical andregulatory experience is available;

• the biosimilar approach is defined by the current recommendations on analyt-ical methods, manufacturing process, and clinical studies conducted for theapproval application;

• by definition, a biosimilar product is not a generic product; subtle differencesbetween biosimilar and reference may exist and call for a prior experiencebefore using them. In order to facilitate a later follow-up (pharmacovigilance),patients receiving a biosimilar must be clearly identified.In the same general recommendations, the same biological reference must be used

for the whole program of comparability of quality safety and efficacy studies, in orderto ensure that responses during the comparability exercise be obtained with a singlecomparator, having all along the studies concerning the same form and same dosage,the same types of impurities and variants linked to its manufacturing process. Abiosimilar’s active substance must be similar in molecular and biological terms to thereference product. For instance an a 2a-interferon cannot be biosimilar to an a 2b-interferon. It is strongly recommended for the biosimilar medicinal product to havethe same form, dosage and route of administration as the reference medicinalproduct. If it is not the case, additional data must be given in the context of com-parability exercise to justify these differences. Any difference between biosimilar andreference must be justified by appropriate study, case by case. A consultation withregulatory authorities is recommended for discussing these approaches.

Quality Control Approach

Biosimilars are biological products developed according to their own manufac-turing process. Scientific data coming from pharmacopeia’ monographs orpublished in the literature on the reference biological medicinal product areconsidered as limited in order to establish the similarity between biosimilar and

From the Biosimilar Concept to the Marketing Authorisation 25

reference at the active substance and finished product levels, for they are notrelevant enough. Only a comparability exercise will allow the evaluation ofsimilarity in terms of quality, safety and efficacy. Based on a complete qualitydossier combined with sensitive analytic tests, the comparability exercise at thequality level allows the reduction of the number of non-clinical and clinicalstudies, compared to a complete approval application file.

A complete quality file, comparable to the file required for the referencemedicinal product approval, is always required for biosimilars approval applica-tions. It is completed with quality, non-clinical and clinical comparability databetween the reference medicine and the biosimilar medicine.

Biosimilars’ Manufacturing Process

The biosimilar is defined by its manufacturing process specific to the activesubstance and to the finished product (as for the reference medicinal product).These processes must be developed and optimized according to current regulatoryrecommendations, covering aspects of the molecular expression system and of theproduction cells, culture, purification, viral protection, formulation excipients,interactions with primary packaging materials, as well as their possible conse-quences upon the finished product characteristics. Besides, every medicine isdefined by its molecular composition, which is itself defined by its manufacturingprocess which introduces its own impurities. For these reasons, the biosimilar isdefined by:• the molecule itself, including variant products and impurities;• the manufacturing process which may play upon molecular characteristics and

impurities.The company that develops the biosimilar must master all these issues in terms

of reproducibility and robustness of the processes involved. It is recommended thatclinical data in the comparability exercise be obtained with the biosimilarmanufactured according to the final manufacturing process that will be used forbatches to be marketed. Otherwise ‘‘bridge’’ studies will be needed.

Quality Comparability Exercise

Quality issues are essential for a biosimilar and their potential impact on safety andefficacy must always be evaluated. A step by step approach is recommended inorder to analyse and justify any difference in the quality attributes betweenbiosimilar and reference. It is not demanded that the quality attributes be identicalas minor structural differences may exist for the active substance, due to thepost-translational modifications’ variability or differences in impurities profile.


These may be acceptable but must be justified, notably in terms of their possibleimpact upon safety and efficacy of the finished product.

Analytical Methods

Characterisation studies must be conducted according to regulatory currentrecommendations concerning the active substance and at the same time the finalproduct to demonstrate that the biosimilar quality is comparable to that of thereference. The analytical methods must be chosen according to the product’scomplexity and must be able to detect differences between biosimilar and refer-ence. The comparison is done with validated analytical methods assessingcomposition, physical properties, primary and higher degree’s molecular structure,different forms related to post-translational modifications, and biological activities.Several biological tests are needed; they use various approaches in order tocompare the biosimilar’s and reference’s biological activity. Activity expressionmust be stated in international units, if an international standard exists.

A biosimilar’s derived products and impurities must be identified and comparedto its reference’s using current available techniques. Stress studies are used toshow specific degradations (i.e. oxidation, dimerization) and accelerated stabilitystudies lead to profiles of stability that can be compared between biosimilar andreference.

Impurities related to the manufacturing process (proteins and DNA [deoxyri-bonucleic acid] of the host cell, reagents, purification impurities) are specific anddepend on the manufacturing process of each product. Because of this thecomparability exercise may not be applied in an absolute manner. However thebiosimilar, as the reference product, must meet the same level of requirementsdescribed in the recommendations on biotechnology-derived products quality.

Specifications

As with any biotechnology-derived product, the specifications are based on aselection of tests depending on the given product. The rationale for fixing thelimits of acceptation criteria must be described and developed following the sameapproach as for any biological medicinal product. Each acceptation criterion mustbe established and its justification must be based on batches used in non-clinicaland clinical studies, on batches produced in a reproducible way, and on datacoming from comparability exercise (quality, safety, efficacy).

To fix specifications, the company that files the marketing authorisationapplication must use a global reasoning: this application is based on experienceacquired from the product being developed and its reference medicine. Data must


show, if possible, that the limits of a given test are not wider than the variabilitydeviations observed with the reference medicinal product.

Conclusion on Quality

The quality aspect in a biosimilar’s development is essential. It is on that aspectthat mostly lays the demonstration of similarity between biosimilar and reference.The quality file of a biosimilar must contain the two following demonstrations:• characterisation and production full studies, on active substance and on finished

product;• a comparability exercise to evaluate the quality and similarity of the biosimilar

and the reference. These studies have to be interpreted in the context of safetyand efficacy comparable between biosimilar and reference. In the biosimilarapproach, if data concerning quality are crucial, they have, however, to becompleted with data coming from non-clinical and clinical comparative studies,more limited than those required for the development of a brand new medicine.

Non Clinical and Clinical Aspects

Quality, safety and efficacy are key issues that must be followed during a medi-cine’s whole life. For a typical chemical medicine, pharmaceutical development iswell-defined. It includes data to document the pharmaceutical quality and iscompleted by preclinical, called toxicological data, before the first humanadministration. Clinical development requires data concerning a proof of concept,dose evaluation and demonstration of efficacy in pivotal studies conducted in themedicine’s target population. Based on quality, preclinical and clinical studies,stored during its development, the medicine may be ready to be filed in order to geta Marketing Authorisation (MA). As for chemical medicines, the applicationbiosimilar approach necessitates the development of the manufacturing process(for the active substance and finished product), and the demonstration of safety andefficacy through non-clinical and clinical studies. However, as the reference bio-logical medicinal product has been already approved and used for many years inthe European Union, its data are available in the public domain. Consequently, abiosimilar development calls for less non-clinical and clinical data than a newmedicine; some of this data may be taken as given with the reference product andbe used as ‘‘support’’ data in the biosimilar file. Thus, if the reference is approvedin several clinical indications, and its mechanism of action is the same in allapproved indications, then it is possible to assume that there is a ‘‘therapeuticsimilarity’’ between reference and biosimilar and to extrapolate the biosimilarefficacy demonstrated for one indication to other indications of the referencemedicinal product.


Preclinical Approach

Preclinical studies are comparative and generally include in vitro studies ofreceptor bindings and tests on cells already found in quality data provided for thebiological activity evaluation. These studies can establish the comparability interms of their mechanism of action between compared products and identifycausality factors in case comparability could not be established. In vivo studies onrelevant animal studies must be added, while taking into account the regulatoryguidelines in force.

Preclinical study has to evaluate, when the animal model allows it:• activity in connection with the pharmacodynamic effect relevant to the clinical

application;• non-clinical toxicity determined with a single and repeated dose; it is not

necessary to have toxic dose finding studies, as they are already known.Measurements in toxicokinetics include the determination of the level ofantibodies with the study of crossed reactions and of the neutralisation capacity;the studies must last long enough to show any difference relevant in terms oftoxicity and/or immune response between the biosimilar and the referenceproduct;

• if necessary, local tolerance comparative studies.Other routine toxicological tests (safety pharmacology, reproductive tests,

mutagenicity, carcinogenicity) are not necessary. The preclinical studies programis a limited program due to the fact that the toxicology data are known for thereference medicinal product and it is not necessary to repeat all the studies to knowthe biosimilar.

Clinical Approach

The exercise of clinical compatibility is done step by step; it generally starts withpharmacokinetics and pharmacodynamics studies in healthy volunteers. Thesestudies are followed by efficacy and safety comparative studies. In most cases, theclinical efficacy studies are conducted to demonstrate a therapeutic equivalencebetween the biosimilar and the reference in a population of patients chosen for themost sensitive to the studied medicinal product effects in order to evidence anydifference that could be exist between biosimilar and reference. However, even ifefficacy is demonstrated through a therapeutic equivalence test, a biosimilar tol-erance may differ from the reference’s if there are differences in terms of qualityattributes not apparent or difficult to analytically demonstrate. These differencesmay have unpredictable clinical consequences, and a biosimilar clinical tolerancemust be continuously evaluated before and after its marketing authorisation.

During the evaluation of clinical tolerance, a special attention has to be paid toimmunogenicity, because patients may develop against the biosimilar as againstany recombinant protein in some circumstances; these antibodies could have


clinical consequences. The immunogenic potential of a biological medicinalproduct differs between products and depends on several factors like the activesubstance’s nature and structure, impurities, excipients of the medicine, manu-facturing process, route of administration, and target population. These differencesmay compromise the product in vivo behaviour, with, as a consequence, unde-sirable effects for the host that may minimize the intended clinical effect withpotentially lethal reactions.

Different approaches based, for instance, upon the response of the epitope toHuman Leucocyte Antigen (HLA) polymorphism, or the immunological responsestudied in relevant animal models, may be used to evaluate a biosimilar immu-nological profile. However, if these responses are important to identify the anti-genic profile, they are not predictive of the immunological response to thebiosimilar in vivo. Evaluation of a biosimilar antigenic profile in patients iscomplex because of the difficult measurement of antibodies’ level (unavailabilityof immune serums, absence of appropriate standards, interference of endogenousproteins, limits of analytical methods, etc.) Similarly, the simple comparison ofproducts of the same therapeutic class, although interesting on a theoretical level,is not enough and may be the source of misinterpretation.

Overall, the decision to put a biosimilar on the market is made if its efficacy issimilar and its immunogenic profile is at least comparable or improved incomparison to the reference product. However, this decision is made on limiteddata. The comparability program may disclose substantial differences in terms ofimmunogenic profiles but is probably unable to detect minor differences and rareevents. For that, clinical trials complemented by a pharmacovigilance program areessential for evaluating a recombinant protein’s safety in patients. Some unde-sirable effects are very rare and require a follow-up during the medicinal product’swhole life; this is particularly true for biosimilars.

Recommendations in Onco-Haematology

Hematopoietic Growth Factor (rG-CSF)

The file of a biosimilar of Recombinant Granulocyte Colony-Stimulating Factor(rG-CSF) that positions itself as similar to a medicine already approved in theEuropean Community and whose patent has expired must demonstrate itscomparability in terms of non-clinical and clinical quality with the reference product.

The human G-CSF is a protein made of 174 amino acids with an O-glycosyl-ation site on a threonine. The recombinant protein obtained in E.coli is not gly-cosylated and presents an additional terminal methionine. The rG-CSF protein hasa free cysteine and two disulfide bonds. The medicines rG-CSF obtained byexpression in E. col (Filgrastim�) and in CHO [Chinese Hamster Ovary](Lenograstim�) are clinically used for several indications:


• reduction of the duration of a severe neutropenia after a cancer chemotherapy ormyelosuppressive treatment followed by a bone marrow transplant;

• mobilisation of hematopoietic stem cells in peripheral blood (Peripheral BloodProgenitor Cell [PBPC]);

• treatment of severe congenital, cyclic or idiopathic neutropenia• treatment of persistent neutropenia in Human Immunodeficiency Virus (HIV)

patients.

Doses Vary with IndicationsG-CSF acts on target-cells through a membrane receptor. Only one soluble isoformthat attaches itself to the extracellular part of the receptor is known. The extra-cellular binding domains of known isoforms are identical. Consequently, G-CSFeffects are mediated by only one class of receptors.

The approval application and marketing of a G-CSF biosimilar requirecomparative studies of non-clinical and clinical quality.

Non Clinical Program for rG-CSF

The non-clinical program includes:• comparative pharmacodynamic studies:

– in vitro at receptor level on adapted cellular models, to measure biologicalactivity;

– in vivo on neutropenic and non neutropenic rodent models, in order tocompare the biosimilar effects to those of the reference;

• toxicology studies with a single or repeat dose to a relevant species for at least28 days.Other routine toxicity tests are not required.

Clinical Program for rG-CSFThe clinical program to compare biosimilar to the reference product includes:• pharmacokinetics studies in crossed single dose for the different routes of

administration (subcutaneous, and intravenous) in healthy volunteers. Studiedparameters include the area under the curve (AUC), the C max and T � with anevaluation performed according to bioequivalence general principles;

• pharmacodynamics studies—the absolute number of neutrophils is the phar-macodynamics marker the most relevant for G-CSF activity. The pharmaco-dynamics study may be done during the pharmacokinetics with a dose selectionin the ascending linear part of the dose–response curve; repeat dose studies maybe necessary. CD34+ level is a secondary pharmacodynamic parameter;

• the clinical model suggested for efficacy clinical studies is the prophylaxis ofsever neutropenia after cytotoxic chemotherapy in a group of patientshomogenous in terms of tumour type and in terms of programmed and validatedchemotherapies according to the tumour stage. A two-arm study comparing


biosimilar and reference is recommended with the measurement of frequencyand duration of neutropenia as the efficacy main criterion. The company mustjustify the clinically acceptable difference in the sever neutropenia duration(ANC \0,5 9 10/L) between the biosimilar and the reference. This evaluationwill be done during the first cycle of chemotherapy;

• G-CFS effects are mediated by only one class receptors and the results ofclinical comparability obtained on the model may be extended to other indi-cations of the reference product;

• clinical safety must be evaluated from a cohort of patients who have receivedrepeat doses of biosimilar, preferably during the comparative phase of theclinical trial. The total exposure of patients must correspond to the normalexposure of the conventional treatment with a corresponding number of che-motherapy cycles. The duration of the study must not be shorter than six monthsand must integrate immunogenicity data. The number of patients must besufficient for evaluating the secondary effects including bone pains and bio-logical parameters;

• a strengthened program of pharmacovigilance must be implemented with a riskmanagement plan. The two must take into account that immogenic events arerare but serious in patients with a chronic administration.

Erythropoietin

Human erythropoietin (EPO) is a 165 amino acid-glycoprotein produced in thekidney, that stimulates the production of red blood cells. The medicine is obtainedfrom recombining DNA technology in mammal cells able to express a glycosyl-ated protein.

The recombinant protein has the same sequence as the natural protein butdiffers by the number and types of isoforms. The protein’s glycosylation influencesefficacy and safety including the protein’s immogenicity.

Erythropoietin based medicines are indicated in various conditions such asanemia in patients suffering from chronic renal insufficiency in patients treated by acancer chemotherapy inducing an anemia, and also in some programs of autologoustransfusions differed in order to increase the number of autologous blood donations.The active substance’s mechanism of action is the same for all indications currentlyapproved but the doses to get the desired response vary a lot and are generally higherfor cancer indications. The medicine is injected by IV or SCD.

As it is generally well-tolerated, EPO allows a range of therapeutic concentrationrelatively wide. The hemoglobin content reached allows a control of the bonemarrow stimulation and consequently of doses and periodicity of the treatment. Thehemoglobin content increase varies considerably between patients and depends onnumerous factors like dose and administration rhythm but also the level of iron in thebody, basal content of hemoglobin and endogenous erythropoietin, and concomitanttreatments or patient’s underlying condition, such as inflammation.


The pharmacodynamic response must be under control to avoid serious unde-sirable effects like high blood pressure and thrombotic complications. Cases ofPure Red Cell Aplasia resulting from the production of anti-erythropoietinneutralising antibodies have been observed, mainly in patients with chronic renalinsufficiency and treated with sc injections. Stemming from the fact that usuallythese antibodies’ production is a very rare event, clinical studies for pre-marketingauthorisation do not identify these events. Other considerations have to be takeninto account for erythropoietin approval applications that are their possibleangiogenic action and tumour promoter. Thus the study population selection isparticularly important.

The approval application files for a new biosimilar erythropoietin involve thedemonstration of comparability with the reference product in terms of quality,safety and efficacy.

Non Clinical Program for EPOThe non-clinical studies include:• pharmacodynamic comparative studies:

– in vitro to evaluate the absence of altered response on receptors, with tests ofbinding to receptors or with cellular proliferation tests. Some tests come fromquality comparative studies;

– in vivo to evaluate the erythrogenic action on relevant animal models. Infor-mation on the erythrogenic activity may be obtained through toxicity studieswith repeat doses or specifically with a methodology like the one described inon mice in the European Pharmacopeia (Normocythaemic Mouse Assay);

• single and repeated dose toxicity studies on a species relevant to rats. Thestudies must last at least four weeks and include a toxicokinetics evaluation;

• local tolerance studies, notably with repeated doses with subcutaneousinjections.Other routine toxicity tests are not required.

Clinical Program for EPOThe clinical program is comparative between copy and reference; it is made ofpharmacokinetics studies in crossed single dose for the different routes ofadministration (subcutaneous, and intravenous) in healthy volunteers. The dosehas to be chosen in the sensitive part of the dose–response curve. Studiedparameters include the area under the curve (AUC), the C max and T �. The bio-equivalence margins must be beforehand defined and justified;• the pharmacodynamic parameters must be preferably studied during pharma-

cokinetics. In single dose studies, the most relevant parameter is the number ofreticulocytes, for it is a pharmacodynamic marker of erythropoietin’s activity.However this marker does not substitute for efficacy, since it is not directlycorrelated with hemoglobin level;

• clinical biosimilarity must be demonstrated by comparative clinical studiespowerful enough, randomised and in parallel groups between biosimilar and


reference. As pharmacokinetics and efficient doses differ between IV and SCroutes, studies must be conducted on each mode of injection. The studies maybe conducted either separately for each route, or for one route with appropriate‘‘bridge’’ data for the other route. Double blind studies are preferable in order toavoid any bias;

• sensitivity to erythropoietin is better for patients who have a deficit in endogenouserythropoietin than for patients without a deficit. Patients with chronic renalinsufficiency without major complications will be preferred as a model popula-tion for the biosimilar’s clinical trials. The other possible anemia causes will beexcluded from the comparability studies. The populations in dialysis and pre-dialysis shall not be mixed, as the doses needed to maintain the hemoglobin levelare not the same;

• it is possible to demonstrate efficacy’s similarity through different options andrecommendations. Two different clinical trials are conducted; the trials maycombine a phase of anemia correction by sc injections (for instance for a pre-dialysis population) and a maintenance phase by iv injections (for instance foran haemodialysis population). During the correction phase, the dynamicresponse and the dose may be determined by carefully checking on the safetyprofile of biosimilar’s patients. This phase may include treatment-na patients orpatients already on treatment after a three-month treatment free period. In themaintenance phase, patients must have an optimal titration on reference productfor at least 3 months. After this period, they are randomised between biosimilarand reference product, while keeping the erythropoietin prerandomisation dose,as well as the periodicity and the administration route. For the correction phase,the responder rate or the change in hemoglobin level may be chosen as aprimary endpoint of clinical activity. Anyway, dosing erythropoietin remainsthe trial’s secondary endpoint. A four-week evaluation period is necessary for astudy lasting 5–6 months, for the correction phase as well as for the mainte-nance phase. The studied must be designed according to a methodology fit forevaluating the equivalence between the two products; another approach is toconduct a comparative efficacy study for one route of administration and toprovide, for the other route, data resulting from ‘‘bridge’’ studies comparativeof PK/PD in single dose and multiple dose, conducted in a population sensitiveto erythropoietin (for instance healthy volunteers). The PK/PD study in multipledoses must last four weeks minimum, with a fixed dose of EPO with a primaryendpoint fixed on the evolution of hemoglobin level; in all cases of immuno-genicity comparative data are required for sc route. In comparative sc routestudies, a total duration of twelve- months’ treatment is required.

• the clinical safety data are generally sufficient to provide a satisfactory data basefor pre-marketing authorisation. The undesirable effects’ follow-up notablyincludes high blood pressure and its possible aggravation and thromboembolicevents. The company must file immunogenicity data coming from a 12 months’period for the biosimilar’s application file. A validated test sensitive to detectingearly and late antibodies must be implemented during correction and mainte-nance phases. Searching for the presence of neutralising antibodies or Pure Red


Cell Aplasia episodes during the pre-authorisation phases is crucial; it must becomplemented by an adequate post-MA follow-up. The data allowing to dem-onstrate a clinical similarity come from the comparative trial on the populationconsidered the most relevant (chronic renal insufficiency), both in iv and in sc (ona number of patients big enough, as it is commonly accepted that the sc route ismore immunogenic that the iv route);

• the typical pharmacovigilance program is completed by a risk management plannotably taking into account rare and serious secondary effects like Pure Red CellAplasia of immune origin and the EPO’s potential action of tumour promoter;

• as EPOs’ mechanism of action is identical for all approved indications for thereference product, and since there is a known EPO receptor, the demonstrationof efficacy and safety in the chronic renal insufficiency population makespossible the extrapolation to other reference medicine’s indications for the sameroute of administration.

Conclusion

The biosimilar approach based on an exercise of comparability with preclinicaland clinical data, in addition to quality data, allow pharmaceutical companies tofile a shortened file (compared to a standard complete file required for a newbiotechnology-derived medicinal product) in order to obtain the MA of a biologicalproduct similar to the reference biological product; it is called a ‘‘biosimilar.’’ It ismore necessary to establish a given level of similarity in terms of quality issues thanin terms of safety and efficacy, for the biosimilar to be approved in one or allindications of the reference medicine’s. Biosimilars are above all biologicalmedicinal products characterized by their own quality profile. The long-term con-sequences of possible differences between biosimilar and reference are not wellknown because the clinical trials, conducted over a short period, are designed todemonstrate the equivalence of efficacy and pharmacodynamics. The long-termsafety profile will be known only after several years of these products’ use. Becauseof that fact, a biological medicinal product cannot be substituted by a biosimilarmedicinal product (as for standard generics) before having collected long-term dataon efficacy and safety of the product in all populations to be treated. Currently, inFrance, the substitution of a biological medicinal product by a pharmacist is notpossible. Only a medical prescription in controlled conditions may allow the sub-stitution of a reference biological medicine by a biosimilar.

Further Reading

• Directive 2004/27/EC du Parlement européen et du conseil modifiant la directive2001/83/EC instituant un code communautaire relatif aux médicaments à usagehumain (31 mars 2004)


• EMEA/CHMP/437/04 Guideline on Similar Biological Medicinal Products(October 2005)

• EMEA/CHMP/BWP/49348/05 Guideline on Similar Biological MedicinalProducts containing Biotechnology-Derived Proteins as Active Substance:Quality Issues (CHMP adopted 22 February 2006)

• EMEA/CPMP/ICH/5721/03 ICH Topic Q5E Comparability of Biotechnologi-cal/Biological Products (CHMP adopted December 2004)

• EMEA/CPMP/ICH/302/95 ICH Topic S6 Step 4 Note for Preclinical SafetyEvaluation of Biotechnology-Derived Products (CHMP adopted September 97)

• EMEA/CHMP/42832/05 Guideline on Similar Biological Medicinal Productscontaining Biotechnology-Derived Proteins as Active Substance: Non-ClinicalAnd Clinical Issues

• EMEA/CHMP/BMWP/31329/05 Annex Guideline on Similar BiologicalMedicinal Products containing Biotechnology-Derived Proteins as ActiveSubstance: Non-Clinical and Clinical Issues. Guidance on Similar MedicinalProducts containing Recombinant Granulocyte-Colony Stimulating Factor(CHMP adopted 22 February 2006)

• EMEA/CHMP/BMWP/94526/05 Annex Guideline on Similar BiologicalMedicinal Products containing Biotechnology-Derived Proteins as ActiveSubstance: Non-Clinical and Clinical Issues. Guidance on Similar MedicinalProducts containing Recombinant Erythropoietins (CHMP adopted 22 mars2006)


Immunogenicity

J.-L. Prugnaud

Introduction

Immunogenicity may be defined as the power of an antibody to induce an immuneresponse in a given individual in appropriate conditions. An antibody may beantigenic without being immunogenic if, at least in certain conditions and incertain subjects, it is able to induce a response in binding itself in a specific mannerto immunoreceptors. Nevertheless, the antibody injection will not trigger in thegiven individual and given conditions an immune response. Thus, in the definitionof immunogenicity there is a quantitative notion linked to circumstances that isfound again in the fact that some antibodies are very much immunogenic andothers are little [1].

All proteins are potentially immunogenic. The therapeutic proteins can there-fore always trigger an immune response when they are injected into the body. Thisimmune response may have more or less serious consequences, from a simpletolerance reaction to antibodies, up to therapeutic inefficiency when the antibodiesare neutralising. Antibodies produced against therapeutic proteins like erythro-poietin (EPO) [2], hematopoietic growth factors (GM-CSF) [3] and thrombopoi-etic/megakaryocyte (TPO/MGDF) [4] may have big consequences to the point ofblocking not only the exogenous protein’s activity but also of the endogenousprotein with the serious complications inherent in these actions.

The production of antibodies against biotechnology-derived proteins likeinsulin, factor VIII or IX, or interferons, does not have the same serious conse-quences and doctors go on with the treatments in presence of the antibodies,adapting the doses of therapeutic protein [5–7].

J.-L. Prugnaud (&)Prugnaud Consulting, 13 ave Jean Aicard, 75011 Paris, Francee-mail: [email protected]


37

The consequences of antibodies produced against monoclonal antibodies havebeen observed since their first use, particularly when these proteins are derivedfrom animal or bacterial proteins. The reactions observed could be of a generalorder such as systemic reactions, during these products’ injection, local reactionsor reactions of acute hypersensitivity (generally not due to the antibodies). Theimmune reactions of anaphylactic type or allergic reactions have been observedrelatively often and are now rarer because of the better purification of proteinsproduced by recombinant DNA technology and the humanisation of proteinskeletons of monoclonal antibodies [8–10]. The production of neutralisingantibodies may correspond to several types of mechanisms like the directbonding to a biological activity site or to a site which is not in direct relationbut impedes its activity by a changed structural conformation. The non-neu-tralising antibodies bind to the therapeutic protein site without affecting thebiological activity site. If they don’t directly neutralise the biological target, theymay change the medicine’s bioavailability by increase of the clearance of thebonding complex made with a result identical to that of biological activityneutralisation [8].

Whatever the nature of the antibodies produced, the immune responses gen-erated by therapeutic proteins pose a significant problem of safety and efficacy forthe authorities in charge of evaluating and approving the marketing of biologicalmedicinal products. Recommendations [11] concerning the evaluation of thebiosimilar’s immunogenic profile comparatively to the references have beenpublished. These recommendations are based on a multifactor approach taking intoaccount the mechanisms involved; the different factors that may be part of theimmune response, and the level of expression of antibodies and the possible rarityof the response observed. It is compulsory for the companies to evaluate thebiosimilar’s immunogenic risk, case by case, to ensure its safe use. No specificmethod is recommended, taking into account the variability and multiplicity offactors involved, but actions to take must be identified before clinical trials start, aswell as evaluations that will be performed during pivotal clinical trials and eval-uations that will be done after marketing of the product, notably within theframework of a risk management plan.

To measure the consequences of the risk linked to immunogenicity during thetherapeutic use of biosimilars, immune mechanisms at play must be examined aswell as factors having an influence on immunogenicity.

Immune Mechanisms

It is currently accepted that antibodies’ formation against therapeutic proteins maybe done by two canals: a typical immune response comparable to the one directedagainst foreign proteins, and a response of tolerance breaking down vis-à-vis the‘‘self-proteins’’ [9].

38 J.-L. Prugnaud

Typical Immune Response

In general, a typical immune response occurs after the administration into the body of aforeign protein. Antibodies are produced according to the following classic sequence:• antigen-presenting cells capture the foreign protein and cleave peptide bonds,

degrading proteins into peptides that associate then to class II molecules of thehistocompatibility major system or HLA system in humans. The lymphocytes TCD4+ recognise the peptides presented by the class II HLA molecules andactivate the lymphocytes B that produce the antibodies specific to these pep-tides derived from the foreign protein. There is an ‘‘immunologic memory’’phenomenon that translates into ‘‘effector memory’’ and by a ‘‘central mem-ory’’. The effector memory leads to a rapid destruction of the foreign proteinwhen it is reintroduced in the body, when the central memory is the capacity forproducing antibodies and effector T cells more rapidly and for a lower dose ofantigen than during the primary response;

• the classic immune response is observed when foreign proteins of animal,microbial or plant origin, are administered to humans. The antibodies formationis usually quick, from a few days to one or several weeks, often after a singleinjection. The antibodies are most often of the neutralising type. They maypersist for a long period.A classic type immune response may be observed with human proteins pro-

duced with the recombinant DNA technique in patients presenting an innateimmune deficit. In fact, the production of neutralising antibodies has beenobserved with the recombinant coagulation factor VIII and with the recombinantgrowth hormone in such patients.

Immunologic Tolerance Breakdown Response

Numerous recombinant human proteins are homologous to endogenous proteinstructures also called ‘‘self-proteins’’. After injection, these recombinant proteinsdon’t induct an immune response, as the body is ‘‘tolerant’’ to these moleculesconsidered as self-proteins. However, antibodies directed against these recombinantproteins can be observed in an immunologic tolerance breakdown. The formation ofantibodies by this process is slow; it often appears in patients who had receivedchronic treatment for months. Generally, these antibodies disappeared when thetreatment ended. The exact mechanism of immunologic tolerance breakdown is notknown. Physiologically, there are lymphocytes B which recognize self-proteins.There lymphocytes B are called ‘‘self-reactive’’. When a self-reactive lymphocyte Binteracts with an epitope present in a repetitive form, a cross-linking of receptors Boccurs, which leads to an activation of self-reactive lymphocyte B, and the synthesisof antibodies. This mechanism could be related to the immune system of lympho-cytes B recognising microbial type patterns, independently from the self/non-selfdiscrimination exercised by these cells. An illustration of this mechanism of

Immunogenicity 39

tolerance immunologic breakdown would be the immune system response directedagainst the repetitive protein structures such as aggregates. Some aggregates dem-onstrate a structure analogous to repetitive structures. This structure would induce aninitial activation of lymphocytes B with production of IgM type antibodies. Theoccurrence of this IgM/IgG transformation, elucidating the IgG synthesis specific ofaggregates, is done according to a mechanism unknown to this day. Some studiessuggest that aggregates are internalised after reacting with the receptors of lym-phocytes B. After internalisation, lymphocytes B would synthesise cytokinescapable of activating other lymphocytes B. Inversely, the authors suggest the exis-tence of a different mechanism involving auxiliary T lymphocytes, called T helpers.However, no published data has described the presence of specific T cells in patientsexhibiting antibodies directed against therapeutic proteins. Finally, the absence ofassociation with HLA haplotypes and the absence of immunological memory sug-gest a mechanism independent from T lymphocytes.

Factors Influencing Immunogenicity

Therapeutic proteins immunogenicity is influenced by different factors. Some con-cern the protein’s very structure, how to produce it; with its purification degree, itsformulation in order to make a medicine out of the therapeutic protein, the treatmenttype and the patients’ characteristics, plus other factors possibly not known.

Structural Factors

Proteins are complex molecules with a primary, secondary and tertiary structure.

Primary StructureChanges within the primary structure may be the cause of an immunogenicreaction. Several cases are well known and published in the literature:• changing an insulin amino acid is enough to lead to a strong immunogenic response,

whereas two amino acids inversion only leads to a pharmacokinetic change;• the homology degree of a recombinant protein with the natural protein may

explain an immunogenic reaction but the well-known case of recombinanthuman a interferon that shows 10–23 amino acids different from human ainterferon (homologous for about 89 %) does not lead to immunogenicityexacerbation. Foreign proteins like streptokinase, salmon calcitonine etc., areknown for inducing classic immunogenic reactions in patients;

• reactions of oxidation or deamidation of amino acids are known for triggering animmunogenic reaction by forming new epitopes. It is the example of humanrecombinant a interferon with one methionine, oxidised because of a modificationof the purification process, that has led to non neutralising antibodies formationand which, returning to the initial process, has stopped being immunogenic;

40 J.-L. Prugnaud

• the modification of stability characteristics of a protein with aggregates for-mation may have significant consequences in terms of immunogenicity bytolerance breakdown of the immune system.

GlycosylationGlycosylation is also an important factor in therapeutic protein immogenicity.Glycosylation is a post-translational modification, cell and species dependant. It iswell-shown that proteins with a human structure produced in eukaryote non-humancells give immunogenic human responses. Similarly, the glycosylation level playsan important role. A b interferon produced in E.coli is more immunogenic than thatproduced by mammal cells since the latter’s glycosylation, helping its solubility,decreases the formation of immunogenic aggregates. A protein’s glycosylation candecrease immunogenicity by masking antigenic sites.

PEGylationSome recombinant proteins are ‘‘pegylated’’ in order to modify their pharmaco-kinetics. PEGylation of a protein is the process of attachment of polyethyleneglycol (PEG) chains to the skeleton of the protein. The result is a decreased totalhalf-life of the protein, proteolytic enzymes protected, and sometimes maskedimmunogenic sites. The new proteins so obtained differ by their conjugatedstructure, their molecular size and their spatial conformation (linear, branched, ormulti-branched chains). Often does PEGylation lower the protein’s immunoge-nicity, probably through multiple mechanisms related to blocked antigenic sites,solubility improved, and lower administration frequency of the therapeutic protein.In general a branched PEG protein is more efficient than a linear PEG proteinbecause of improved immunological properties. However, examples have beenpublished in the literature of PEG proteins more immunogenic than non-PEG(PEG-rhMGDF and rh-TNF coupled with PEG dimer).

Secondary and Tertiary Structures

The significance of protein spatial conformation is well-known for its biologicalactivity as well as tits stability. Partial modification of spatial conformation mayoccur after shear, by shaking, or by temperature modification (for example: tem-perature rise or freeze/thaw cycles). Aggregation phenomena may expose newepitopes to the protein’s surface for which the immune system is intolerant. Thatleads to a standard immune response.

In other conditions, protein aggregation may lead to presenting a multimericantibody, which is known for not triggering B lymphocyte tolerance breakdown.This is why, in a therapeutic protein’s analysis, it is important to look for thepresence of aggregates and to limit their presence to a low level in the formulatedmedicine.

Immunogenicity 41

Impurities and Other Production Contaminants

Therapeutic proteins obtained through recombinant DNA technology are producedin various cellular systems where production-linked protein impurities originate.These proteins are called host cell proteins (HCP), considered ‘‘non-self’’ proteinsby the immune system and may lead to antibody formation by the standardimmune mechanism. If by nature anti-HCP antibodies don’t neutralize the bio-logical activity of the therapeutic protein of interest, they can nevertheless haveconsequences in terms of general effects including skin reactions, allergies, ana-phylaxis or a serum sickness. Other contaminants, such as impurities, coming fromchromatographic column resins or from enzymes used for refining therapeuticproteins’ purification, may be found as traces in the finished product. Someimpurities may be released by some compounds used for the capping process.These impurities may play the role of amplifier for the immune response, even ifthey are not able to initiate an immune response themselves.

Producing therapeutic proteins is a complex process that, with time, necessi-tates change, sometimes major, to keep the production system at an even level ofefficacy. It is important to verify that the same levels of quality and safety aremaintained. Several examples exist on the modification of a recombinant protein’simmunogenicity with time, involving production bacterial cells endotoxin level.DNA G-C patterns from bacteria or degraded proteins are able to activate Toll-like(Toll-like receptors are a class of proteins playing a key-role in the innate immuneresponse. They are transmembrane proteins containing receptors that detect dangersignals located in the extracellular milieu, a transmembrane medium, and anintracellular medium allowing the activation signal transduction) receptors and actas adjuvants. However, the action of these impurities is limited to non-humanproteins with a pseudo-vaccination activity. The adjuvants are unable to stimulatean immune response based on a T lymphocyte’s response independent of Blymphocytes’ tolerance breakdown.

Manufacturing Process and Formulation of the Medicinal Product

The finished product’s formulation and conservation conditions are important tomaintain the therapeutic protein’s biological activity and stability.

Two particularly interesting cases could illustrate how important formulation andconservation conditions are. The first case concerns erythropoietin (EPO). Cases ofPure Red Cell Aplasia (PRCA) after treatment by EPO are known, but rare. NicoleCasedevall [2] has shown the incidence of anti-EPO antibodies’ formation, exoge-nous as well as endogenous, after administration of recombinant human EPO(rHuEPO). These cases have occurred after a changed formulation of the finishedproduct, with human albumin used as a stabiliser replaced by polysorbate 80. Dif-ferent hypotheses to explain the immune system tolerance breakdown after admin-istration of rHuEPO lead, among other things, to the impurities’ extraction from thesyringes plunger rod stopper, playing ‘‘booster’’ in presence of EPO [12]. During the

42 J.-L. Prugnaud

analysis of batches called into question, no increase of the level of aggregates or ofthe level of truncated or degraded EPO has been evidenced. In that case, there musthave been several factors having fostered the immune system’s tolerance break-down. In particular, the subcutaneous route of administration may be incriminated, asit is known to produce a pseudo-vaccination effect.

The second case involves the conservation conditions of a freeze-dried for-mulation of interferon a-2a (rHuINF a-2a) that has been stabilised by humanalbumin. At room temperature, a partial oxidation of rHuINF a-2a has made easierthe formation of aggregates with intact interferon and albumin. These aggregatesled to the therapeutic preparation’s immunogenicity [13].

These cases illustrate how important the finished product formulation study andthe evaluation that has to be done of the possible consequences of a changecompared to the initial formulation or its conservation conditions are. It is alsoparticularly important to rigorously analyse the levels of impurities issued from thetherapeutic protein’s system of production.

Patients and Subsequent Treatment Factors

Patient characteristics, as is the case with their genetic statute and type of disease,are known to influence the response and type of immune responses. It is well-known that patients suffering from severe haemophilia with less than one percentof factor VIII, with time, develop inhibitors to the administration of anti-haemo-philic factors of plasmatic origin or derived from recombinant DNA technology.The most plausible explanation resides in the absence of recognition of coagula-tion factors by the immune system as human proteins of the ‘‘self’’ [5].

In the case of EPO above, with its formulation change, only patients with achronic renal insufficiency presented an immune system breakdown. Cancerpatients treated for their anemia by rHuEPO did not present this secondary effect.This illustrates the conditions promoting an immune tolerance breakdown:• chronic treatment with repeated doses for months, even years;• absence of concomitant immunosuppressant treatment;• route of administration (the subcutaneous injection is more immunogenic than

intra-muscular, itself more immunogenic than intra-venous injection).

Case of Monoclonal Antibodies

The first monoclonal antibodies, of mouse origin and obtained as early as 1975,were produced either by ascite’ fluid or hybridoma technology. OKT3 used in thekidney cancer of total mouse origin has demonstrated its immunogenicity after itfirst administration by the standard route of immune system as protein of ‘‘non-self.’’ The anti-monoclonal antibodies were neutralising and blocked any repetitiveulterior use.

Immunogenicity 43

Technological developments, notably applying the recombinant DNA tech-nology to the production of monoclonal antibodies, has led to the conservation ofnon-human active variable parts and their graft on the constant humanised parts ofimmunoglobulin. In that, in this way, have been obtained monoclonal antibodiescalled chimeric, then humanised, and finally totally human. Today, completelyhumanised antibodies are obtained through developing technologies such as thephage display, meaning presentation of peptides on the surface of filamentousphages or in transgenic animals.

Humanised monoclonal antibodies are less or very little immunogenic, even ifat a very low level there persists a possible induction of antibodies. Treatments bymonoclonal antibodies generally require an injection of large and repetitive dosesin the range of several hundreds of mg. These doses must be examined with therequired level of impurities, notably as far as aggregates are concerned. Even ifpurification leads to a level lower than 0.5 % of contaminant aggregates, theinjected amounts are without commune measure with the other recombinanttherapeutic proteins for which the quantity of injected protein matter is in the rangeof one microgram or one hundred nanograms. Real attention must be paid in thatcontext. It is difficult (based on results of published or conducted studies in thecourse of therapeutic trials), to closely follow the level of formed antibodies. Infact, the level of circulating and persistent monoclonal antibodies may mask theformation of induced antibodies. Concomitant treatments of patients by cancermedicinal products acting on the immune system or treatments by immunosup-pressants may reduce antibodies’ neo-formation for these patients. It remains thatattention must be paid to the potential immunogenicity of monoclonal antibodies,for they got properties that contribute to it [14]. They can, by themselves, activateT lymphocytes and complement pathways. It has been shown that the loss ofglycosyled N-linked chains of the Fc part of immunoglobulin, (as it reduces theactivity of the Fc function), may lead to a reduced immunogenicity.

The next authorisation of monoclonal antibody biosimilars will require thattheir immunogenicity to be particularly checked. This will imply that a similaritydemonstration at the quality level be especially well-studied during comparabilitystudies at the level of active substance and finished product. These studies willsurely have to be supplemented by more complete safety comparative studies.

Conclusion

Numerous factors influence therapeutic proteins’ immunogenicity. At this date it isnot possible to completely predict therapeutic protein immunogenicity before theimplementation of clinical trials. Immunogenicity is an event that may generallyoccur with therapeutic proteins. Clinical consequences may vary. The presence ofaggregates or the formation of aggregates in the formulation of a therapeutic proteinis one of major factors known for increasing immunogenicity. It has been shown thatchanges in the manufacturing process and in the finished product formulation con-tribute to modifying the preparation’s immunogenicity. If the consequences of these

44 J.-L. Prugnaud

changes are difficult to predict, in vitro tests and tests on immunocompetent trans-genic mice are being developed and could lead to an evaluation before clinical trials.These tests do not give any information completely predictive of the therapeuticprotein’s immunogenicity, but may allow a comparison a formulation to the other, acopy to its reference.

In particular formulation and sources of production of biosimilars, medicinalproducts should be explored. If regulatorily speaking, the formulation must beidentical to the reference product, production cells and purification techniques willalways be different. Due to this fact, contaminants coming from the productionsystem will always be different. Thus the biosimilar’s immunogenicity in thera-peutic condition must be particularly well explored. Most therapeutic proteinsinduce antibodies in a small number of patients. Post-MA monitoring is especiallycrucial. Consequently European regulatory authorities ask for a risk managementplan to be put in place by pharmaceutical companies once they’ve got the mar-keting authorisation.

Publications on immunogenicity cases have shown how important it is to ensurethe traceability of batches produced and administered to patients in order to be ableto more easily locate the incriminated product, and, if needed, batches in which theevents could have taken place. Traceability is not only for biosimilars but for alltherapeutic proteins that the patient may receive in case interchangeability may beaccepted.

References

1. Dubois M (2008) L’immunogénicité des protéines thérapeutiques dans Développement detechniques analytiques pour l’évaluation des protéines théra-peutiques et des biomarqueurs parspectrométrie de masse, Thèse de Doctorat, Université Pierre et Marie Curie Paris VI. p 106

2. Casadevall N et al (2002) Pure red-cell aplasia and antierythropoietin antibodies in patientstreated with recombinant erythropoietin. N Engl J Med 346(7):469–475

3. Gribben JG et al (1990) Development of antibodies to unprotected glycosylation sites onrecombinant human GMCSF. Lancet 335(8687):434–437

4. Li J et al (2001) Thrombocytopenia caused by the development of antibodies tothrombopoietin. Blood 98(12):3241–3248

5. Jacquemin MG, Saint-Remy JM (1998) Factor VIII immunogenicity. Haemo-philia 4(4):552–5576. Chance RE, Root MA, Galloway JA (1976) The immunogenicity of insulin preparations.

Acta Endocrinol Suppl (Copenh) 205:185–1987. Fakharzadeh SS, Kazazian HH Jr (2000) Correlation between factor VIII geno-type and

inhibitor development in hemophilia A. Semin Thromb Hemost 26:167–1718. Shankar Gopi (2007) A risk-based bioanalytical strategy for the assessment of antibody

immune responses against biological drugs. Nat Biotechnol 25(5):555–5619. Schellekens H, Jiskoot W (2007) Immunogenicity of therapeutic proteins. In: Pharmaceutical

biotechnology: fundamentals and applications, 3rd edn. Informa Healthcare, New York p 12510. Crommelin Daan JA (2007) Immunogenicity of therapeutic proteins. In: Handbook of

pharmaceutical biotechnology. Wiley, The Netherlands p 81611. EMEA/CHMP/BMWP/14327/(2006) Guideline on immunogenicity assessment of

biotechnology-derived therapeutic proteins

Immunogenicity 45

12. Sharma B et al (2004) Technical investigations into the cause of the increased incidence ofantibody-mediated pure red cell aplasia associated with Eprex�. Eur J Hosp Pharm 5:86–91

13. Hochuli E (1997) Interferon Immunogenicity: technical evaluation of interfer-on-alpha 2a.J Interferon Cytokine Res 17(Suppl 1):S15–S21

14. Schellekens H (2002) Immunogenicity of therapeutic proteins: clinical implications andfuture prospects. Clin Ther 24(11):1720–1740

46 J.-L. Prugnaud

Substitution and Interchangeability

J.-L. Prugnaud

Introduction

The new arrival of the copy of an original medicinal product on the market inevitablyposes the questions about the substitution and interchangeability of products. Onehas to clearly define what the concepts of exchanging a medicine for another are, andwho does the exchange, and under what conditions that exchange may be done.

Substitution of Generics and of Biosimilars

In France, substituting an original medicine by a copy is ruled legal for genericmedicinal products in the Public Health Code and in the Social Security Code. Thesubstitution is based on the principle that a pharmacist may legally substitute anoriginal medicine whose patent has fallen in the public domain by a generic—copyor the original product—if the latter is listed in generics groups and if theprescribing doctor does not formally oppose in writing on the prescription, to thissubstitution. Thus the substitution has a regulatory status at a national level. It isthe possibility that is given to the pharmacist to replace the medicine corre-sponding to a brand name by another medicine differently named (but whoseactive substance, strength and pharmaceutical form are identical), so as to ensurethe patient with the same treatment. A generic medicinal product marketingauthorisation is given after review of an approval application file made of acomplete pharmaceutical part and a clinical part on bioequivalence, showing that

J.-L. Prugnaud (&)Prugnaud consulting, 13 ave Jean Aicard, 75011 Paris, Francee-mail: [email protected]


47

the generic product has the same bioavailability as the medicine to which it iscompared.

The substitution concept is thus attached, in France, to positive lists of generics.Substitution implies their own conditions of marketing, their inclusion on a list ofgenerics and the prescribing doctor’s non-opposability.

What About Biosimilar Medicinal Products?

The definition of a biosimilar medicinal product is very precise (see chapter‘‘From the biosimilar concept to the MA’’) and states that if it does not fill theconditions of definition of a generic medicine, notably because of ‘‘differences linkedto raw material or differences between manufacturing processes of the referencebiological product,’’ preclinical and clinical trials results must be provided for itsapproval. Consequently, a biosimilar medicinal product approval application file isdifferent from a generic medicine approval application file. The Code of PublicHealth refines this approach, stating that biosimilars are not to be included in lists ofgenerics. Stemming from that fact, biosimilar medicines cannot be substituted by thepharmacist. This approach is French and subject to national regulation. Althoughbiosimilars’ MA is European and obtained through a centralised procedure, substi-tution is an approach which varies country by country, according to rules that reg-ulate, among others, medical coverage through social protection.

Interchangeability: Suggested Definition

In the Code of Public Health, it is not said that biosimilar medicinal products arenot interchangeable; that is to say that a reference medicine cannot be exchangedfor a biosimilar. In fact, nothing legally forbids the interchangeability of anoriginal biological medicinal product by a similar biological medicinal product incompliance with MA’s indications. But this exchange falls under a medical act ofprescription under the sole responsibility of the attending doctor. From this, adefinition of interchangeability may be described as ‘‘the possibility, by a medicalprescription, to exchange an original medicine for a copy and vice versa’’. Thisconcept of interchangeability as so defined (that is a ‘‘possible exchange’’) will bebetter understood in European countries as well as in other regions of the worldthan the concept of substitution that implicitly or explicitly implies a regulatoryoverlook according to the concerned country.

Biosimilars’ Interchangeability and Conditionsto be Implemented

Are biosimilar medicinal products interchangeable and what conditions have to bemet for an optimum change?

48 J.-L. Prugnaud

http://dx.doi.org/10.1007/978-2-8178-0336-4_2

All biosimilars that have to this date been granted a European marketingauthorisation are copies of proteins derived from recombinant DNA technology.Moreover, the marketing authorisation of these biological medicinal products issupplemented in France by particular conditions of prescription and dispensingfalling under the statute either of hospital reserve or initial hospital prescription.Some medicinal products reserved for hospital use may however be resoldaccording to outpatient needs by hospital pharmacists. When they are available inthe pharmacies, they are dispensed under the pharmacist’s responsibility. In allcases and whatever their dispensing statute, they are not substitutable by thepharmacist. The exchange is decided by the doctor by way of his prescription.

In the interchangeability framework, must biosimilar medicinal productprescription be reserved only for doctors defined in the reference medicinalproduct statute? The answer has been legally given: biosimilar medicinal productsfollow the same prescription rules as reference products that they copy.

Generics may be prescribed by their common name (INN International Non-proprietary Name), given by WHO (World Health Organisation) to the medicinalproduct’s active substance. Is this possibility transposable to biosimilar medicinalproducts? As it has been explained in previous chapters, due to the complexity ofbiological products and the demonstration of similarity-only (and not of strictequivalence between active substance and finished product) differences, if onlyminimal, may exist at the complex molecular structure level that, in most cases, isnot represented by the sole chemical name attributed by the WHO to the activesubstance. Currently it appears that only the INN is not representative enough ofthe biological medicinal products (for instance for differences in glycosylation oron other post-translational modifications in direct contact with the producing cell);consequently, the prescription cannot be made by the INN alone. For reasons ofprescription accuracy, patient safety and follow-up quality, it is preferable that theprescription be made with the medicinal product’s brand name or under a namethat makes the production pharmaceutical company identifiable. Suggesting anevolution for biological medicines’ INNs is needed; these suggestions will comefrom adequate WHO studies.

Biosimilars and their reference products are registered in agreement with acentralised European Community procedure. Hence, all Summaries of ProductCharacteristics (SPC) are the same in all EC countries. Concerning generics, SPCsare identical between generics and reference medicinal products, except when, forexample, a clinical indication is still protected by a patent. Could and should it bethe same for biosimilars?

As it has been detailed, biosimilar filing is done through results of compara-bility tests with the reference product, in preclinical and clinical trials. It thereforeseems normal that the information based on these comparative studies assessingsafety and efficacy be transcribed in the appropriate sections of the SPC. Thisangle adds another difference with generics. But it is crucial for a quality infor-mation representative of the medicine for the prescribing practitioner.

Substitution and Interchangeability 49

The possibility of interchangeability must be coupled with the patient’s rigor-ous treatment monitoring and notably with the patient’s exact treatment. The riskof multiple products taken all along a treatment imposes traceability necessary forthe biosimilar as well as for the reference medicine. It could possibly be facilitatedby the new European bar code that gives at the same time the medicine’s name butalso its batch number and its expiry date. Based on this information, it should beeasier to follow, batch per batch, the medicinal products administered to thepatient. Today this traceability is not required by the authorities, unlike the specialcase of blood derived medicines. It is strongly desirable in order to more easilyknow treatments’ chronology and, in case of possible undesirable effects, thistraceability will give a better grip on these events’ history and chronology.

Interchangeability Practices

If substitution by the pharmacist is not possible, what about implementingbiosimilars’ interchangeability?

A patient’s treatment choice is the attending doctor’s responsibility. Medicalprescription, in all cases, must give with precision the name of the medicine, itsstrength (notably in case of several strengths), doses and duration of treatment. Fora generic, the prescription may be expressed in INN. For a biosimilar, thepractitioner must mention with precision the medicine’s name or its chosenequivalent because of the pharmacist’s forbidden substitution and other compul-sory elements in the prescription. The change of medicine is made by the attendingdoctor according to criteria linked to the particular patient monitored. Dependingon the status of dispensing and prescription defined for the medicine marketing, itis possible that the medicine could be changed only by the doctor who had initiatedthe treatment, and not by the patient referring attending doctor.

Interchangeability management inside hospitals falls under particularconditions. The selection of medicines for the hospital drug formulary is made by amedico-pharmaceutical panel, the committee for medicinal products and sterilemedicinal devices (COMEDIMS).

This panel builds, among other things, the policy enabling the choice of allmedical products listed in the drug formulary. Biosimilars’ arrival onto thecompetitive market of medicines imposes a policy of selection for products thatare not identical to reference products but only similar, that have no genericsstatute and that have the same indications, either totally or partially. We repeat thatfor biosimilars the pharmacist has no right of substitution and that the change isunder the prescribing doctor’s responsibility.

To make treatment’ interchangeability possible without patients taking risks, achange policy has to be defined within the medicinal products committee. Thispolicy must rest on the following criteria:• the committee puts in place medical recommendations to manage the change;

they concern interchangeable products, their conditions for prescription, their

50 J.-L. Prugnaud

equivalence, the modalities in which the change can be made, the patient’sparticulars, patients for whom the change is possible, patients for whom thechange is not desirable or doable (population at risk, population possibly notstudied in clinical trials, etc.);

• a procedure describes the change’s modalities (drug’s prescription, dispensingand administration players, distribution chain, specific validations by the doctoror pharmacist if needed, specific collection of monitoring data; notablyaccording to the risk management plan, to which may the biosimilar besubmitted, etc.).Such a policy makes knowing the biosimilar product totally necessary in its

pharmaceutical parameters (formulation, strengths, composition) as well aspharmacological and clinical. Not only does the drug’s SPC gives this information,but the European Public Assessment Report (EPAR) supplements the scientificinformation, as it is a document concerning scientific data issued by EMA(European Medicine Agency). Other information is available for COMEDIMS’drafting of recommendations, notably data published in the literature. Initiallythose will concern the reference product and not its biosimilar.

The choice made by the committee may be based on the following criteria:• data provided by EPAR, that have shown the similarity;• SPCs data;• the disease(s) that the medicine addresses;• treatment’s chronicity, doses, and periodicity;• routes of administration—criterion relevant to tolerance and immunogenicity;• pharmaceutical form and possible differences between the biosimilar and

original product;• pediatric data, if necessary;• existence of a risk management plan and its implementation;• number of medicines and their pharmaceutical forms;• evaluation that can be conducted of the potential risks due to interchangeability;• competitiveness of hospital market;• availability of the medicine in town pharmacies;• frequency and duration of tenders;• price and/or cost for a treatment.

Interchangeability is dealt with ‘‘case by case’’. The committee’s selection musttake that into account.

May a hospital have on its drug formulary only one type of treatment—thereferent medicine or the biosimilar medicine?

This question deserves to be asked for it impacts the policy of allocations ofmedicines’ tenders managed by hospital pharmacists. Taking into accountconsiderations linked to substitution and the special conditions of interchange-ability implementation, it is necessary to have a flexible supply of referentmedicinal products and their biosimilars in order to fill no substitutable medicalprescriptions. All solutions may be considered from the listing on the drugformulary with or without physical inventory, until available at the wholesalerdistributor’s of references that have not been chosen for the drug formulary and not

Substitution and Interchangeability 51

directly stored in the hospital. Whatever the solution chosen, it takes into accountthe specificities of that type of drug: non-substitutable biological medicinalproducts, possibly interchangeable, under the responsibility of the prescribingdoctor who has initiated the prescription and compliance with it.

Conclusion

The arrival of biosimilars on the competitive market of biomedicines draws theplayers involved in the potential use of these products to engage in a reflection onthe concept of their substitution and interchangeability. If we cannot stand back,the lack of knowledge and use of these medicinal products may be a temporaryfactor in a wait-and-see approach of their prescription. It is possible to define, caseby case, as for the drafting of development and registration programs aimed athealth authorities, recommendations and modalities of use that will frame thedispensation of these products to patients. These will guarantee the safety andgood use of biosimilars.

Further Reading

• Directive 2003/63/EC Definition of Biological product (Part I—3.2.1.1.b)• Directive 2004/27/EC art. 10(4) Biologics: Similar Biological Medicinal

Product• Lekkerkerker F (2009) Are there safety concerns for biosimilars? International

Journal of Risk and Safety in Medicine 21 (2009): 47–52• Revers L, MA, Phil D, Furczon E, HBSc, MBiotech (2010) An introduction to

biologics and biosimilars. Part II. Subsequent entry biologics: Biosame orbiodifferent ? Canadian Pharmacists Journal 143 (4): 184–91 (July 2010)

• Simoens S (2008) Interchangeability of off-patent medicines: a pharmacoeco-nomic perspective. Expert Rev. Pharmacoeconomics Outcomes Res 8(6): 519–21

• Substitution des génériques. Loi n8 2007-248 (art. 8) 26 février 2007• WHO/BS/09.2110 (2009) Guidelines on Evaluation of Similar Biotherapeutic

Products

52 J.-L. Prugnaud

G-CSFs: Onco-Hematologist’s Pointof View

D. Kamioner

Introduction

Febrile Neutropenia (FN) is associated with an important rate of morbidity andmortality at a high cost for society. FN is still a major threat for patients under-going cancer chemotherapy, leading to a loss in quality of life. An increase ofmortality that can reach 9.5 % appears after hospitalisation for FN. Several riskfactors have been identified in order to evaluate the individual FN risk. Thesefactors are patient-linked: age, general health, but also underlying disease(extension, co-morbidity) as well as the chemotherapy protocol used. In order toprevent FN induced by chemotherapy, an antibioprophylaxis and the prescriptionof Granulocyte colony-stimulating factor (G-CSF) have been somewhat success-fully used.

Neutropenia (NP) and FN may lead to a late administration and/or a reduceddose of chemotherapy, hence influence over the disease’s evolution.

For Nicole M. Kurderer, the global hospital mortality is 9.5 %; patients withoutmajor co-morbidity have a mortality risk of 2.6 %, when a major co-morbidity riskis associated with a 10.3 % risk and anything more than a major co-morbidity isassociated with a mortality risk of at least 21.4 %.

The cost to society is high: the average hospital stay is 11.5 days and theaverage cost is 19 110 dollars per FN episode; patients hospitalised for more than10 days (35 % of patients) represent 78 % of total cost.

D. Kamioner (&)Service de cancérologie et d’hématologie, Hôpital Privé de l’Ouest Parisien,14 avenue Castiglione del Lago, 78190 Trappes, Francee-mail: [email protected]


53

The major mortality risk factors for hospitalised patients are fungal infection,BG infections, pneumonia and other lung diseases, brain, kidney and liverpathologies (Table 1).

Hemopoiesis takes place at the bone marrow level, notably at the axial skeletonlevel and long bones. The purpose of using G-CSF is to mobilize bone marrowstem cells and promote precursors proliferation and differentiation.

Overview

Today three products are available in France: filgrastim (Neupogen�), lenograstim(Granocyte�), and pegfilgrastim (Neulasta�), whose mechanisms of action arerecognised as follows:

Filgrastim, r-metHuG-CSF

Filgrastim is a human recombinant factor-stimulating granulocyte colonies pro-duced by DNA recombinant technique based on an Escherichia coli strain (K12).

It is indicated in the reduction of neutropenia duration and of incidences offebrile neutropenia in patients treated by cytotoxic chemotherapy for a malignantdisease (except for chronic myeloid leukaemia and myelodysplastic syndromes), inthe reduction of neutropenia duration for patients receiving a myelosuppressanttreatment followed by a marrow graft and having an increased risk of severeprolonged neutropenia, and in mobilisation of peripheral stem cells in circulatingblood.

Long term administration of filgrastim is indicated for patients, children, oradults alike, suffering from severe congenital, cyclic, or idiopathic neutropenia,with a level of neutrophils B0.5 9 109/l and a history of severe or recurrentinfections, to increase the neutrophil count and reduce infectious episodes’ inci-dence and duration.

Table 1 Protocols ofchemotherapy associatedwith a risk of NF [20 %

Breast cancer Bronchial cancer LMNH

AC/Docetaxel 5–25 ACE 24–57 SCC DHAP 48

Paclitaxel AC 40 Topotecan 28 SCC ESHAP 30–64Doxo/Docetaxel33–48

Doce/Carbo 26 NSCC CHOP 21 17–50

Doxo/paclitaxel 21–32 VP/CDDP 54 NSCC

TAC 21–24

NSCC, Non small cell carcinoma; SCC, small cell carcinoma

54 D. Kamioner

Finally filgrastim is indicated in the treatment of persisting neutropenias(neutrophil count B1 9 109/l) in patients infected by HIV at an advanced stage inorder to reduce the risk of bacterial infection when the other options aimed atcorrecting neutropenia are inadequate.

Pharmacokinetics: there is a positive linear correlation between the dose offilgrastim administered by SC or IV injection and the serum concentration. AfterSC administration at recommended doses, the serum concentrations of filgrastimare maintained above 10 ng/ml for 8–16 h.

Lenograstim, rHu G-CSFIt is produced by a recombinant DNA technique on Chinese hamster ovarian cellsand is indicated for reducing the duration of neutropenia in patients (with non-myeloid neoplastie) receiving a myelosuppressive treatment, followed by a bonemarrow graft and having an increased risk of severe and prolonged neutropenia,and in the reduction of severe neutropenia duration and complications associatedin patients during well-known chemotherapies, known to be associated with asignificant incidence of febrile neutropenia and mobilisation of hematopoietic stemcells in peripheral blood. The safe use of lenograstim has not been shown whenused with cancer agents with cumulative or predominant platelet lines’ (nitro-sourea, mitomycin) myelotoxicity. In these situations, using lenograstim couldeven lead to an increased toxicity, especially for platelets.

Pharmacokinetics: lenograstim is a factor stimulating neutrophil progenitors; asit has been demonstrated by the increased number of CFU-S and CFU-GM inperipheral blood. It notably increases the neutrophil count in peripheral blood,within 24 h after administration. This increased neutrophil level is dose-dependenton between 1 and 10 lg/kg/day. The use of lenograstim in patients who receive abone marrow transplant or are treated by cytotoxic chemotherapy significantlyreduces the neutropenia duration and associated complications.

PegfilgrastimPegfilgrastim, like filgrastim, is produced by recombinant DNA technique from anEscherichia coli strain (K12). The filgrastim produced that way and purified fromE.coli, then goes through a chemical modification in order to introduce PEG(polyethylene glycol) chains on the molecule. These PEG residues are aimed atslowing down filgrastim’s degradation and hence at increasing pegfilgrastim’shalf-life in comparison with filgrastim’s.

It is indicated in the reduction of neutropenia’s duration and incidence of febrileneutropenia in patients treated by cytotoxic chemotherapy for a malignant syn-drome (except for chronic myeloid leukaemia or myelodisplastic syndromes).

Pharmacokinetics: after a single SC administration of pegfilgrastim, the serumconcentration peak appears between 16 and 120 h after injection and serumconcentration stays stable during the neutropenia period following myelosup-pressive chemotherapy. Pegfilgrastim elimination is not a linear function of thedose; pegfilgrastim serum clearance decreases when doses are higher. As the

G-CSFs: Onco-Hematologist’s Point of View 55

clearance is self-regulated, pegfilgrastim’ serum concentration quickly declines asearly as the beginning of the restoration of the neutrophil levels.

Limited data show that pegfilgrastim’s pharmacokinetic parameters are notmodified in patients older than 65.

Biosimilars

The expiration of patents protecting a whole bunch of biological medicinalproducts these last years has led, in Europe, to approval application and in somecases to marketing of medicines called biosimilars.

If chemical synthesis is the basis of ‘‘simple’’ molecules, biotechnology’s benefitresides in the higher capacity of cells to produce complex molecules such as, forinstance, human proteins. Therapeutic proteins have three-dimensional structures ofhigh complexity (see chapter ‘‘Biosimilars’ Characteristics’’). Only a precise con-figuration of these structures leads to an interaction sufficient enough with receptorsand therefore, at the end, their biological action. A modest rise in temperature may,for example, make the protein go to another three-dimensionalstate, which in returnmay lead to a loss of biological function and an increased immunogenicity.

One can therefore understand why clinical studies are necessary to demonstratethe equivalence between biosimilar and reference medicine in therapeutic condi-tions. Guidelines of EMEA (European Medicines Agency) define general require-ments in terms of quality as well as clinical studies; they require approval applicationdata similar to those of the reference medicine. Requirements differ according theconcerned protein (EPO [erythropoietin], G-CSF, etc.) and are defined in specificannexes (http://www.emea.europa.eu/ema/pdfs/human/biosimilar for G-CSF).

While a pharmacokinetic bioequivalence proof is sufficient for generics’authorisation, this authorisation is granted to a biosimilar only on the basis of moreextended clinical studies demonstrating a therapeutic equivalence with the refer-ence medicinal product.

Sixteen therapeutic areas are concern: haematology (11 %) and cancer (7 %)represent 18 % of therapeutic areas concerned by biosimilars; add infectiousdiseases (19 %) and these three specialties represent 37 % (see EMEA andAFSSAPS 2005). The market is very wide if France only is considered, it reaches8.9 % for G-CSF (however almost three times less than EPO).

Prerequisite of EMA Guideline on Similar Biological MedicinalProducts Containing Recombinant G-CSF

Pharmacodynamy StudyThe number of neutrophils is a pharmacodynamic marker that depends on G-CSFactivity. The pharmacodynamic effect of product tests and reference products mustbe compared in healthy volunteers.

56 D. Kamioner

http://dx.doi.org/10.1007/978-3-642-15320-4_1

http://www.emea.europa.eu/pdfs/human/biosimilar/

Clinical Efficacy StudiesSevere neutropenia subsequent to cytotoxic chemotherapy prophylaxis in ahomogenous group of patients is the recommended clinical model in order todemonstrate the comparability of the test product with the reference product.

Alternatively, models including pharmacodynamic studies on healthy volun-teers may be used in order to demonstrate comparability.

Recommendations of EMA (G-CSF Biosimilars Annexes)

A study conducted on healthy volunteers is a model more sensitive for evaluatingrG-CSF efficacy than a study in patients undergoing chemotherapy, becausehealthy volunteers’ bone marrow (contrasting with the marrow insufficiencypatients’) wholly responds to G-CSF treatment.

CHMP (Committee for Medicinal Products for Human Use) has issued positiveopinions for filgrastim’s biosimilar products for neutropenia’s treatment sinceFebruary 2008: these biosimilar versions are similar to Neupogen�, a referenceproduct.

Besides original G-CSF, biosimilars have recently showed up: Tevagrastim�,Ratiograstim�, Zarzio�, Filgrastim Hospira�, Filgrastim Hexal�, and FilgrastimMepha� (Suisse). Some are already on the market, others will be soon.

As written above, the arrival of an original medicine’s copy on the marketinevitably poses the questions of substitution and interchangeability of productsamong themselves.

Has to be defined and what is implied:• the concept of exchanging a medicine with another;• who does the exchange;• in what conditions that exchange may be performed.

Although the MA of recombinant proteins called biosimilar is European andgranted by a centralised procedure, substitution is an approach that varies from onecountry to another, according to the rules that concern, among others, the coveringof health care expenses in the context of social protection.

However, in the Code of Public Health, it is not said that biosimilar medicinalproducts are not interchangeable. In fact, nothing legally forbids the inter-changeability of an original biological medicinal product with a similar biologicalmedicinal product in compliance with MA’s indications. But this exchange fallsunder a medical act of prescription under the sole responsibility of the attendingdoctor. From that, a definition of interchangeability may be given as ‘‘the possi-bility, by a medical prescription, to exchange an original medicine for a copy andvice versa.’’ (see chapter ‘‘Substitution and interchangeability’’).

Numerous questions are raised for substitution, interchangeability, and eventoxicity of biosimilars.

About G-CSF:• the products currently on the market are biosimilars of figrastim, not lenogra-

stim: in fact these two products’ MAs sensibly differ;


• a pegligrastim’s biosimilar will soon be on the market;• filgrastim’s potential toxicity is not identical to lenograstim’s.

As for any medicine, long-term undesirable effects may occur in the case ofsystematic substitution and the monitoring will be difficult, if not correctly tracedin the patient’s file.

Without entering into polemics that followed EPOs’ wide use, G-CSF leuke-mogenic risk must be checked as well as, bone pains, rise of CA 15/3, etc.

Because of some structure variations inherent to the molecule complexity, fil-grastim’s biosimilars (like all biosimilars) cannot be strictly identical to the ori-ginal product. The current limited clinical experience requires a special vigilanceduring the exchange of an innovative product by a biosimilar and their prescrip-tions’ very rigorous follow-up.

A Key Point is the Prescription, in Practice, of Biosimilars: UnderWhat Name, INN (International Nonproprietary Name) or BrandName?

If for generics the prescription expressed in INN facilitates the pharmacist sub-stitution, it can’t be the same for biosimilars, for which the pharmacist substitutionis not authorised in France, and in many other European countries. The INN aloneis not and has never been sufficient to determine the medicine’s prescription anddispensing. Let’s remember that the prescription and dispensing of medicines (andnaturally of biosimilars) is defined by an adequate work group of the Afssaps(French medicinal products agency). It is this agency that has recommended allprovisions in terms of forbidden substitutions for biosimilars. It is also this agencythat took the provisions of prescription and dispensing for biosimilars identical tothose of the original products (hospital initial prescription, hospital reserve),resale, availability in retail pharmacies, etc.). The recommendations that we maycurrently give are based not only on precautionary principle, but on a principle ofsafety and a possible efficient traceability of the medicine at the patient’s file leveland at the medicine’s dispensing by the pharmacist. Today G-CSFs have to beprescribed under their brand names or the INN followed by the producer’s name tofacilitate the pharmacist’s task and to be sure that the prescription be the bestpossible description of the product that will be administered to the patient.

Is the Same Interchangeability Possible and/or Reasonablefor all G-CSFs’ Indications?

As mentioned above, the requirements given in the guidelines on the marketingauthorisation filing of G-CSF’s biosimilars specify that the clinical model chosen forthe comparability of biosimilar and reference is the prophylaxis of severe neutropeniainduced by cytotoxic chemotherapy in a homogenous group of patients. Because of asame mechanism of action of G-CSFs in the various MA indications of reference

58 D. Kamioner

products, the biosimilar’s producer may ask for all indications’ extension, based on thedemonstration of clinical comparability in the recommended model. The EuropeanMedicines Agency’s MAs are not to be put in question and if all indications have beengranted, it is because the biosimilar responds to a positive benefit/risk ratio.

However, the observation duration of clinical trials that have been conducted mustbe watched. Is this observation duration long enough, does it cover all possibilities toobserve all undesirable effects and notably the tolerance specific to immunity? Theresponse is partly given by authorities themselves who ask for a risk managementplan in postmarketing of G-CSF biosimilars. This plan is destined to cover theinsufficient knowledge of these products’ tolerance for longer use.

For the prescribing doctor, the attitude is (in the early days of a biosimilar’savailability of reserving his prescription of the medicinal products to first-timeaccessing patients, and then with the publication of results of risk managementplans) to widen the medicines’ interchangeability. Such an attitude is not overlycautious; it guarantees a better safety for the patient.

Could Situations Occur When Interchangeability Becomes NonApplicable or When, at Least, The Product Could be Changedwith Precautions?

As mentioned above, only a longer experience on medicines’ tolerance, through a risksmanagement plan, will adequately help dealing with the issue of interchangeability.

The specific situation of multiple changes supposedly induced by nomadicpatients has to be taken into consideration. The prescribing doctor must have accessto the patient’s whole file with the exact prescription that the patient has received, thetreatment’s duration, the number of chemotherapy cycles, and thus the number ofG-CSF cycles performed. Immune system involvement depends, amongst otherthings, on the treatment’s periodicity, its length and how it is repeated.

Clinical trials conducted on biosimilars give information on peripheral stem cellmobilisation in healthy volunteers. Have pharmacology/pharmacodynamics studiesbeen conducted on enough healthy volunteers to provide complete information onthe biosimilar’s tolerance? If so, using a biosimilar is no riskier than a new medicine.

The same prudent attitude on changing the medicine is to be considered forcongenital, severe cyclic, and idiopathic neutropenia indications, because of thelong and repetitive treatment. In all cases re-evaluating the long term tolerancebased on risks management plan results is a must.

Could an Hospital Pharmacist Impose the Exclusive useof a Biosimilar, in Agreement with the Hospital Tender?

Today it still looks unlikely, due to insufficient experience and a lack of doctor’straining in the use of biosimilars. The attitude recommended above has to be debated


with the hospital pharmacist and may be validated by the local medicine commissionin which doctors and pharmacists discuss the hospital’s therapeutic choices.

Conclusion

Due to biotechnologies’ development, drug prescription habits change; biosimilarsare a new therapeutic approach to which one will have to adapt.

Special precautions for use involve:• strict compliance with the MA (Marketing Authorisation)• for the prescribing doctor: a possible change of product for different prescrip-

tions while, however, it is not logical to change medicines during a treatmentwhen there is no proof of therapeutic identity.

Further Reading

1. Aapro M, Crawford J, Kamioner D (2010) Prophylaxis of chemotherapy-induced febrileneutropenia with granulocyte colony-stimulating factors: where are we now? Support CareCancer 18(5):529–41

2. Aapro MS et al (2006) EORTC Guidelines for the use of granulocyte-colony stimulatingfactor to reduce the incidence of chemotherapy-induced febrile neutropenia in adult patientswith lymphomas and solid tumours. Eur J Cancer

3. Beney J Les Biosimilaires ne sont pas des génériques. Institut Central des HôpitauxValaisans, Sion vol. 1 n 6, août 2009

4. Dictionnaire Vidal (2010)5. EMEA/CHMP/BMWP/31329/2005 Guideline on Similar Medicinal Products containing

Recombinant G-CSF6. Kamioner D (2008) Les indications des facteurs de croissances leucocytaires. Oncolologie,

Springer 10(5)7. M. Kuderer N et al (2006) Cancer 106(10)8. Möll F (2008) Les biopharmaceutiques et les biosimilaires. Transport, stockage, préparation

et utilisation. pharmaJournal 19:5–89. NCCN (2010) Practice guidelines in oncology Myeloid growth factors vol. 1

10. R Jared A et al (2006) When the risk of Febrile Neutropenia Is 20 %, prophylactic colony-stimulating factor use Is clinically effective, but Is It cost-effective? JCO 24 (19):2975–2977

11. Repetto L et al. (2003) EORTC cancer in the elderly task force guidelines for the use ofcolony-stimulating factors in elderly patients with cancer. Cancer 39:2264–72

12. Schellekens H (2009) Biosimilar therapeutics-what do we need to consider? NDT Plus2(Suppl1):i27–i36 doi:10.1093/ndtplus/sfn177

13. Smith TJ et al (2006) ASCO update of recommendations for the use of white blood cellgrowth factors: evidence-based clinical practice guideline. JCO 24(19)

60 D. Kamioner

http://dx.doi.org/10.1093/ndtplus/sfn177

The Oncologist’s Point of View

C. Chouaı̈d

Introduction

Anemia affects multiple targets, which explains the strong impact of anemia onpatients’ quality of life (Table 1).

Anemia also impacts patients’ survival: in a meta-analysis of 60 clinical trialscorrelating survival and anemia, it has been observed that the risk relative to deathwas 65 % higher [IC 95 %: 54–77] (2) in anaemic patients than in non-anaemicpatients. In short, anemia linked to cancers appear to be frequent, to have multiplefactors, and to have significant consequences. They are a factor in patients’ fatigueand create a tumour hypoxia. They possibly could alter treatments’ response andtheir long-term consequences are not well-known. Thanks to existing medicines,they don’t have to be undertreated anymore.

In oncology the main drugs that are involved for treating anemia are iron byoral administration, or preferably in injections, erythropoiesis stimulating agents(ESA), and red cells transfusions. Cancer patients’ anemia is under-treated.According to ECAS study (European Cancer Anemia Survey) in breast cancer,conducted on 15,000 patients and 1,000 investigators in 24 countries, only 28 % ofpatients have been treated for their anemia:• 7 % were receiving an iron supplementation (average dose 11.7 g/l at initiation);• 12 % an ESA (erythropoietin);• and 7 % were receiving blood transfusions (average dose 9 g/dl at transfusion time);• 74 % of patients, with a hemoglobin level of less than 12 g/dl, were receiving

no treatment for their anemia.

C. Chouaïd (&)Service de pneumologie, Hôpital Saint-Antoine, 184, Rue du FaubourgSaint-Antoine, 75571 Paris, Cedex, Francee-mail: [email protected]


61

Erythropoiesis Stimulating

Before 1988, blood transfusion was the treatment for mild to severe anemia.Natural human erythropoietin (hEPO) is mainly produced in the kidney. EPO actsdirectly on progenitor blood cells located in bone marrow to control erythrocytes’proliferation, differentiation and maturation. It is a 165 amino acid glycoprotein,whose biological activity and pharmacokinetics strongly depend on glycosylation.In 1988, epoetin a (Eprex�) was the first ESA to be approved in the treatment ofchronic renal insufficiency anemia. In 1993 only it was approved for the treatmentof chemotherapy-induced anemia, in Europe. In 1997, epoetin b (Neorecormon�)was approved for epoetin a0 same indications. These two ESAs are glycoproteinsproduced by recombinant DNA technology (rhEPO [recombinant Human Eryth-ropoietin]). They have a primary sequence similar in amino acids to hEPO’s. Theydiffer by the number of isoforms representative of their glycosylation profile. TherhEPOs currently on the market have the following main indications:• anemia of chronic renal insufficiency patients dialysed or not;• chemotherapy-induced anemia of cancer patients;• programmed autologous transfusion.

rhEPOs’ mechanism of action is the same for all currently approved indicationsbut required doses strongly differ according to indications, with doses much higherin cancer indications.

In 2001 an ESA analogue of rhEPOs was approved, but with a longer duration ofaction than darbepoetin a (Aranesp�). It is derived from a modified glycosylationwith an increased content in sialic acid, hence different pharmacokinetic properties.

Table 1 Anemia’s impactCentral nervous system (CNS) Renal function

Cognitive function Reduced perfusion

Mood Water retention

Cardio-vascular system Digestive system

Tachycardia Irregular transit

Weakness

Cardio-respiratory system Genital tract

Effort dyspnea Menstrual disorders

Dyspnea Lower libido

Cardiac decompensation Impotence

Skin Immune system

Reduced perfusion Immunodeficiency

Pallor

Coldness

62 C. Chouaı̈d

Lifespan is prolonged, making an injection possible every one or every three weeksin compliance with the SPC (Summary of the Product Characteristics).

Since 2007, five copies of epoetin a (Eprex�) have been granted a MarketingAuthorisation by the EMA. Filed with biosimilars’ dossiers, these rhEPOs have beenconsidered as similar in terms of quality, safety and efficacy to the reference product.

In 2010 has been licensed an epoetin theta (rhEPO-h) whose structure is relatedto rhEPOs a and b, with minor glycosylation differences. Their biological activ-ities as well as its duration of action are similar to rhEPO b, but this drug has notbeen approved under the biosimilar’s statute, because of differences linked toglycosylation. It is authorised in the other rhEPO’s indications.

Erythropoietins in Anemia Treatments

ESA’s value has been demonstrated on the reduction of transfusion needs and onthe quality of life. In a number of studies, transfusion need is significantly reduced(about 20 %) compared to the group not receiving ESA. This result has beenconfirmed for darbepoetin a in a European study conducted on 705 patients withhemoglobin (Hb) lower than 10 g/dl.

A clear improvement in the quality of life parameters, maximum between 11and 12 g/dl Hb, is shown in the questionnaires follow-up, taking into accountdifferent quality of life aspects, notably fatigue.

Although there are no comparative studies between various rhEPOs, it does notseem that there is a difference in efficacy and impact between epoetins (alpha, betaor zeta) or darbepoetin alpha. On the other hand, rhEPOs’ biosimilars marketingauthorisations give a response in terms of comparison with the reference product towhich they are compared.

In anemia treatments, there are benefits and risks (Table 2).

Table 2 Benefits and risks of treatments by EPO and blood transfusions

ESA Transfusion

Benefits Hb level improved and maintainedSymptoms improvedTransfusion needs reducedNo need for veinous port of entryNo administration in the hospital

Rapid improvement of Hblevel and hematocritSymptoms quickly improved

Risks Thrombo-embolic riskSmall proportion of non respondingpatients potential survival reductionin cancer patients not treated bychemotherapyRisk of tumoral progression notproved in some cancers

Reaction to transfusionCongestive heart failureTemporarily improved HB levelRisk of Iron overloadRisk of viral contamination(Hepatitis B [1:250,000];Hepatitis C [1:2,000,000]; HIV[1:2,000,000])Development of multipleallo-antibodies

The Oncologist’s Point of View 63

If the benefit/risk ratio of anemia treatments by ESA is positive for the licensingauthorities, it has to be noted that a certain number of risks factors exist anddeserve to be known. Thus the population to be treated must be defined, based on afew parameters and when a new treatment must be initiated.

When and for Which Patients can rhEPO Treatmentmay be Prescribed?

Several publications had the impact on survival of treatments by EPO as anobjective. A literature analysis identifies eight studies (out of 59) of good qualityand dealing with that theme. They involve 3014 patients. Table 3 presents theseeight studies. Overall, it seems that there is a risk of high mortality (in the contextof anemia with or without chemotherapy and/or radiotherapy) but essentially whenthe targeted levels to be reached are above 23–14 g/dl.

These analyses’ results must be completed by those of meta-analyses takinginto account a larger number of patients. They are presented in Table 4, with thenumber of studies analysed, the number of patients evaluated, the relative risk ofdeath and the confidence interval at 95 % of the relative risk.

In Bohlius et al. meta-analysis, EPOs significantly decrease transfusion needsfor patients with a level of Hb B 12 g/dl at treatment initiation, and increase thehaematological response defined by an increased Hb level by at least 2 g/dlcompared to the initial level. There is a positive effect of EPOs on quality of lifeparameters. However, the risk of thromboembolic complications is higher inpatients receiving EPO in comparison with those who do not receive it, as there isa higher risk of high blood pressure in patients treated with EPO.

Definitive conclusions cannot be drawn concerning EPOs’ effect on a localtumor response and/or patients’ global survival.

In Bennett et al. meta-analysis [4], the risk of thromboembolic complicationswas increased in patients receiving EPO versus the ones who did not receive it andthe mortality risk was significantly increased in patients receiving EPO versusthose who did not receive it.

In the new meta-analysis of Bohlius et al. [5], treatments by EPO plus trans-fusion have been compared to transfusion only.

With an intention to treat analysis done by independent statisticians who havetaken into account the meta-analysis’ fixed and random effects, overall, EPOsseem to have increased patients’ mortality during the studies’ active phase andaggravated the patients’ global survival (Table 5). However, there is heterogeneityin the analysed studies:• patients with a low basal level of hematocrit had a higher risk of mortality;• patients with a history of thromboembolic complications had a lower risk of

mortality.That is not the case if the data stemming from these studies concern only

patients undergoing chemotherapy.

64 C. Chouaı̈d

If these meta-analyses analyse reliable data, however, they show limits ofinterpretation as for survival data:• profiles and characteristics of patients studied are heterogenous (comorbidity,

tumor stages…);

Table 3 Effect of an EPO treatment on global survival

Study/type ofcancer (n)

Hemoglobintarget

Primary endpoint Results

CHEMOTHERAPY

1. Leyland-Jones(BEST)Metastatic breastcancer(n = 939)

12–14 g/dl 12 month globalsurvival

Decreased global survival: 70 versus76 %, p = 0.01

2. Hedenus(Amgen 161)Lymphoidtumour(n = 344)

13–15 g/dl(M) 13–14 g/dl (F)

Proportion ofresponding patients

Decreased global survivalRelative risk = 1.37, p = 0.04

3. PrepareEarly breastcancer(n = 733)

12.5–13 g/dl Level without relapseand global survival

Level without relapse and globalsurvival decreased global survivaltumor progression accelerationdeaths: 10 versus 14 %

4. Thomas(GOG-191)Cervical cancer(n = 114)

12–14 g/dl Progression freesurvival, global survival,locoregional control

Decreased global survival: 75 versus61 % Survie sans progressiondiminuée: 65 versus 58 %

RADIOTHERAPY

5. Henke(ENHANCE)Head and neckcancer(n = 351)

5 = 15 g/dl(H)5 = 14 g/dl(F)

Global survival,locoregional control

Decreased global survivalRelative risk = 1.38, p = 0.02Decreased locoregional controlRelative risk = 1.69, p = 0.007

6. DAHANCA-10Head and neckCancer(n = 522)

14–15.5 g/dl Locoregional control Locoregional controlRelative risk = 1.44, p = 0.03

WITHOUT RADIOTHERAPY NOR CHEMOTHERAPY

7. WrightNon small celllung cancer(n = 70)

12–14 g/dlQuality oflife

Decreased global survivalRelative risk = 1.84, p = 0.04

8. Smith(Amgen 103)

12–13 g/dl Incidence of transfusion Decreased global survivalRelative risk = 1.3, p = 0.08


• target haemoglobin levels differ from one study to the other. The target levelswere too high because [12 g/dl, going up to 16 g/dl;

• in some studies patients were not receiving any chemotherapy.Observing in some studies a tumor progression under an EPO treatment could

be explained by:• target levels of Hb [ 12 g/dl, not allowing a tumor hypoxia and affecting the

disease progression [6];• re-expression by some tumor types of EPO receptor on the surface of tumor

cells. These seem to respond by a proliferation after stimulation by EPO [7];• EPO off-label use in patients without chemotherapy.

Following another meta-analysis of 12 randomised trials and 2301 patients, ifthe EPO treatments comply with the criteria for use of EORTC (European orga-nisation of cancer research and treatment), then in that case there is no impact onsurvival, tumor progression or mortality by thromboembolic events.

Updating the Guidelines (ASCO/ASH)

Results of these various analyses have led learned societies to offer new recom-mendations for cancer treatments by EPO (Table 6).

Complying with these recommendations must lead to an efficient anemiatreatment in cancer patients and so avoid numerous undesirable effects, includingthromboembolic complications.

In conclusion, treatment by EPO is an etiological treatment of symptomaticanemia that must be adapted to the clinical situation. The EPO treatment has to be

Table 4 Data from meta-analyses

Number of studiesanalysed

Number ofpatients

Death relativerisk

95 % CI

Bohlius et al. 2006[1]

42 8167 1.08 0.99–1.18

Bennett et al. 2008[2]

51 [13122 1.10 1.01–1.20

Bohlius et al. 2009[3]

53 13933 1.17 1.06–1.30

Table 5 Meta-analysis, Bohlius et al. 2009

Population Mortality Global survival

RR 95 % Cl P RR 95 % Cl P

All cancer patients (n = 13933) 1.17 1.06–1.30 0.002 1.06 1.00–1.12 0.05

Studies with chemotherapy (n = 10441) 1.10 0.98–1.24 0.12 1.04 0.97–1.11 0.26

66 C. Chouaı̈d

initiated if Hb level is around\10 g/dl and according to clinical state if Hb level issituated between 10 and 11.5 g/dl, with doses calculated on the basis of weight.EPO should be interrupted in case of absence of response in 6–8 weeks or if Hblevel increase is lower than 1 g in 1 month and without any impact on the recourseto transfusion. EPO should also be stopped if the response is too rapid (increase ofHb level [ 1 g in 2 weeks). Hb level target is 12 g/dl.

Biosimilars

Since 2007, five copies of epoetin a (Table 7) have been granted a marketingauthorisation by the European Medicine Agency. Introduced with biosimilarsapplication files, these rhEPOs have a protein structure similar to rhEPO a- (samenumber of amino acids, same primary and tertiary structure, same number of N andO-glycosylation). All are produced on Chinese hamster ovary (CHO) mammalcells, able to glycosylase the protein structure. Let’s remember that this glyco-sylation is necessary for protein biological activity. Because of the genetic designof production cellular systems specific to each product, there are minor differencesat glycosylation levels, which have no incidence upon the pharmacokinetics ofeach biosimilar comparatively to the reference product (Eprex�). They are allshort action duration ESAs. The analytical comparison gives a similar biologicalactivity.

Active substances’ INNs (International non-proprietary names) are selected bythe WHO and not by European authorities. One of the active substances has aname different from the reference substance. The request for this designation hasbeen made by the producer to ensure a better traceability for the product.

Table 6 Learned societies recommendations for EPO treatment in cancer patients before andafter 2008

Before February 2008 Since February 2008

Therapeuticindication

Treatment of symptomatic anemia of an adult cancer patient in chemotherapy for amalignant non myeloid disease

Initial Hb =s 11 g/dl =s 10 g/dl

Targetvalues

Not specified 10–12 g/dl

Not over 13 g/dl 12 g/dl

Treatment Treatment interrupted ifHb [ 13 g/dl

Treatment interrupted if Hb [ 13 g/dl

Doseadjustment

Reduced dose to maintainHb at its original level

Reduced dose to ensure that the adequate minimaldose is used to maintain Hb at a level fit to controlanemia’s symptoms

References H&N, BEST H&N, BEST, AoC Meta-analyses


However, there are minor differences in the protein chain glycosylations’minority forms; they have no consequences upon pharmacokinetics nor upon thebiological activity of the biosimilar compared to the reference product. For otherbiosimilars, the manufacturers have chosen to keep the same INN for the activesubstance; even in the cases of very minor differences in the protein chain gly-cosylations’ minority forms.

Biosimilars have been approved in compliance with the centralised Europeanprocedure ‘‘of biological medicinal products similar to a reference biologicalmedicinal product.’’ The files filed at the EMEA comprised a complete qualitymodule and shortened preclinical and clinical modules. All studies conducted interms of quality, safety and efficacy were comparative to the reference product.The benefit/risk ratio has been judged positive by the European authorities and hasled to granting a marketing authorisation for the introduced medicines.

Considering that the mechanism of action of erythropoietin was the same in allindications of the reference medicinal product, European authorities have grantedto biosimilars all indications of the reference product based on studies conductedon the relevant model of chronic renal insufficiency patients. However, non-comparative tolerance studies at higher doses used for cancer patients have beenconducted, especially for thromboembolic events’ follow-up in these patients.These studies have shown these biosimilars’ equivalent short term tolerance forcancer patients. For cancer, IV and SC administrations have been accepted.Additional studies have been requested for SC route, for chronic renal insuffi-ciency patients, when the studies have not been conducted because of a contra-indication of this route at the time of the study phases.

Benefiting from a centralised approval application, as the reference products,biosimilars have the same Summary of Product Characteristics in the EU 27 countries.

In general, undesirable effects observed during biosimilars’ clinical trials havebeen comparable to those observed with reference EPO. Three consequent risksare generally associated with EPO treatments:

Table 7 Biosimilars authorised in the EU

Biosimilars References

Brand name ActiveSubstance

Producer Nom ActiveSubstance

Producer

Binocrit� Epoetin a Sandoz Eprex� Epoetin a Jansen-Cilag

Abseamed� Epoetin a MediceArzneimittelPütter

Eprex� Epoetin a Jansen-Cilag

EpoetineAlpha Hexal�

Epoetin a Hexal AG Eprex� Epoetin a Jansen-Cilag

Retacrit� Epoetin zeta Hospira Eprex� Epoetin a Jansen-Cilag

Silapo� Epoetin zeta StadaArzneimittel AG

Eprex� Epoetin a Jansen-Cilag

68 C. Chouaı̈d

• erythroblastopenia (PRCA Pure red cell aplasia) events;• vascular thromboembolic events;• potential risks of tumor progression.

Post-marketing risk management plans have been introduced to follow in cohortstudies the incidences in PRCA; especially in patients suffering from renalinsufficiency treated for their anemia. Monitoring of thromboembolic events and oftumour progression has been carried out in additional post-marketing pharmaco-vigilance plans. Risk minimising plans have also been put into place with mentionin the SPC of a treatment contraindication by a biosimilar EPO for patients whohad developed PRCA in the aftermath of an EPO treatment. In the SPC, attentionis drawn on the possibility to develop PRCA when on EPO. As for thromboem-bolic events, it is mentioned in the SPC that the target Hb level is 12 g/dl. Finally,the potential risk of tumour progression is mentioned in SPC’s ad hoc sections.

Two PRCA cases have been observed during clinical trials with ESAs withdetection of anti-epoetin antibodies. As the investigations of the cause of thesePRCAs are ongoing, European authorities consider it important that current EPOtreatments’ background be monitored and recorded with brand name and/or INNassociated with the producer’s name. For all ESAs it is also recommended that theinformation on the product comprises the mention of keeping the recording ofpatients’ treatments.

What About the Possible Substitution and Interchangeabilityfor EPO’s Biosimilars?

Biosimilars’ substitution and interchangeability are outside European authorities’competence, leaving it up to regional appreciation; that is to say to EU membercountries, the rules to put in place. In France the substitution by a pharmacist of areference product is not possible, for these products are not registrable on thegeneric drugs repertory. This is the very definition of biosimilars.

However, interchangeability of an original medicinal product by a biosimilar ispossible for doctors with regard to their profession’s freedom to prescribe. It has tobe done within the Marketing Authorisation conditions, and especially the con-traindications and precautions for use.

Four ideas may guide the doctor in his approach of biosimilars’ prescription andinterchangeability:• the potential immunogenicity of proteins is known, biosimilars, as well as

reference products cannot avoid the risk of developing in some patients thisundesirable event. With EPOs, the risk of developing antibodies neutralisingboth exogenous and endogenous erythropoietin is a major risk for the patient. Apatient’s follow-up is therefore especially important and any decrease in hae-moglobin levels during a treatment that has been efficient before must trigger asearch for antibodies and guide the decision to interrupt the treatment. It isimportant to draw attention to the patient’s medicine related background,


particularly for cancer patients with hepatitis C treated by interferon andribavirin in whom the PRCA risk is increased;

• targetting 12 g/dl must be imperative for cancer patients treated by EPO, andhelp controlling thromboembolic events. Speed and figures of recovery of Hblevel must be monitored. Recommendations in terms of figures and durationguide the treatment’ interruption (see infra);

• the prescription may be done only with a brand name or in INN if it is followedby the manufacturer’s name. It make the treatment’s traceability easier andmore secure;

• prescribing a biosimilar for a patient treated for the first time relates to anoriginal drug prescription.In all cases, with biosimilars arrival on the biomedicines market, traceability for

all products is crucial. It is recommended by European authorities and it ismentioned in the information made available for doctors and patients. This givesan opportunity to recall that any change in the medicinal treatment of a patientimplies that this treatment has to be explained to the patient and that all infor-mation concerning the medicinal product safe use must be provided.

Conclusion

Five biosimilars of EPO have been authorised by the European Union afterdemonstration of their similarity with an epoetin a (Eprex�) as reference productin all quality, safety and efficiency aspects. Plans of risk management and mini-misation have been put in place to supplement routine pharmacovigilance andbetter inform doctors and patients on use and precautions to take with theseproducts.

Prescribing EPO biosimilars for cancer patients treated for the first timerespond to the same level of precaution and of compliance with the MA as fororiginal drugs. Interchangeability is a medical prescription act that must take intoaccount a number of precautions to ensure the patient’s safety. Substitution by thepharmacist is not possible. European authorities recommend the traceability ofbiosimilars’ as well as that of original medicines.

References

1. Groopman JE et al (1999) J Natl Cancer Inst 91:1616–16342. Caro JJ et al (2001) Cancer 91:2214–22213. Bohlius J et al (2006) J Natl Cancer Inst 98(10):708–7144. Bennett CL et al (2008) JAMA 299(8):914–9245. Bohlius J et al (2009) Lancet 373(9674):1532–15426. Newland, Black (2008) Ann Pharmacother 42:1865–18707. Henke et al (2006) J Clin Oncol 24:4708–4713

70 C. Chouaı̈d

Biosimilars: Challenges Raised byBiosimilars: Who is Responsible forCost and Risk Management?

F. Megerlin

Introduction

The advent of ‘‘biosimilars’’ in the European Union has given three types of hopes:hopes of generic producers, eyeing an important market; hopes of payers (insur-ance companies, hospitals, patients), expecting significant savings; and hopes ofdoctors and pharmacists, hoping that competition will stimulate research for ever-improving treatments (second generation of biological medicinal products). Thischapter gives an overlook on some key points on management in the context ofEuropean biosimilars and also sketches the American approach. Biosimilarity is aconcept subject to different regulatory definitions internationally and to differentdemands for comparability. Due to standard differences across the world, theconcept of ‘‘biosimilar’’ has hence to be considered rigorously within its properregulatory context of market approval (MA).

General Information on Cost Management

Because regulatory, technical and financial barriers to market entry are substantial,the EU experience is still limited. To date only fourteen biosimilar products havebeen approved representing only three active substances (two recombinant growthhormones, five erythropoietins, and seven GCSFs). Three applications have beenwithdrawn (for biosimilar insulins), and one has been rejected for inadequatesimilarity to the reference product (interferon alpha-2a). Only large genericmanufacturers or biopharmaceutical companies with highly developedbio-production skills and the resources necessary for long-term investments can

F. Megerlin (&)LIRAES, Université Paris Descartes, 4 avenue de l’Observatoire, 75006, Paris, Francee-mail: [email protected]


71

consider entering the EU biosimilar market. After having been authorised for the‘‘single’’ European market, biological medicinal products and their biosimilarshave their modalities of use determined by member states. Reimbursement, pric-ing, rules on prescription and dispensing, etc. fall in their remits, not in the EU one.Hence each State has its own specificity, as we have seen with substitution (seechapter ‘‘Substitution and interchangeability’’).

Despite these differences, some challenges are common to all EU memberstates, hospital facilities, physicians, pharmacists and payers. The first challenge isthat a biosimilar’s purchase price is only one component of its total cost of use.

Savings Linked to the Purchase of Biosimilars

What Percentage of Price Reduction is Due to Competition?Expiring patents leads to competition between producers and give hope for lowerprices. For chemical drugs, price reduction may reach more than 80 % of originalproduct’s price. As a result there is a strong price competition as the copy isdeemed ‘‘identical’’ to the original product (that facilitates market penetration).For biosimilars in Europe, the average reduction is about 10–30 % of the referencebiological medicinal product (for retail price; up to 80 % discount for hospitalprices for some products like EPOs, see below).

How to Explain Less Reduced Prices?Bioproduction explains the difference: it requires a high level of industrial know-how and has irreducible costs, still very high—in contrast with the manufacturingcost of small chemical entities that are easily characterizable and identicallyreproducible (see chapter ‘‘Biologicals’s characteristics’’). A biosimilar’s devel-opment costs are about 80–120 million dollars, as opposed to a generic develop-ment cost of 0.4–2 million dollars. This gap is also explained by MA procedures:they are much more demanding, therefore more costly than for generics (about 0.8million dollars) (see foreword). Market penetration is also more costly due toexplanations called for by a new concept (‘‘biosimilarity’’) and by the requiredprecautions. This levelling up explains the limited number of producers qualified tomanufacture biosimilars marketed in the EU, contrasting with the myriad of genericproducers. To date four companies have been master-applicants (producers), whileall others are only co marketing applicants.

Is this Smaller Price Reduction Discouraging?No, the price reduction compared to the original medicinal product is, for sure,smaller for biosimilars than for generics. But the percentage of reduction applies to asensibly higher price. For instance, 30 % of price difference for an annual treatmentof $50,000 represents a yearly savings of almost $17,000. The potential savings aretherefore appealing to buyers (insurance companies, hospitals, patients more or lesscovered, depending on their country), while the benefit remains significant for thecompeting producers. Thus the market is also widely animated by price competition,

72 F. Megerlin

http://dx.doi.org/10.1007/978-2-8178-0336-4_4

http://dx.doi.org/10.1007/978-2-8178-0336-4_1

even if price ranges are smaller. Besides, financial constraints to come for manyprivate and public payers are a powerful thinking incentive.

Strategies Played by CompetitorsThe response varies depending whether that the medicinal product retail prices are fixedor not by public pricing authorities, that they are included or not in health care packagesinvoiced by hospitals, etc. That depends on national health systems and backgrounds ofusage. In general, innovators have a strategy of lowering the prices of their originatorproducts when the market opens to competitors. That sometimes reduces interest forbiosimilar producers but also limits the interest of potential buyers such as hospitals: ontop of the purchase price, they also bear the cost of use or of first year(s) of use: (seebelow). Besides, when there is a negotiation between buyer and producer, price dif-ference between a biomedicine and its biosimilar is not necessarily decisive. Accordingto the national systems, the buyer have to consider the risk that the producerof originatorproducts to whom he is not loyal will make commercial conditions less favourable onother products, etc. These anticompetition practices are illegal but they exist. In somecountries, the fact that competition is possible leads to tenders, etc. That depends onhospital purchase regulation. This complex management will determine in part thehospital costs of health care providing, and consequently its financial balance.

Why is the Competition for Getting Hospital Markets so Strong?Depending on the country, biologics may be dispensed in hospitals only, or need aninitial hospital prescription, etc. In the first case, hospitals are the only market. In thelatter case, hospitals are the key access to outpatient market, which is much moreprofitable. Retail prices in the EU are generally ‘‘listed prices’’ and non negotiable.Competition for primo prescription is thus fierce; hospital prescriptions oftendetermine subsequent outpatient prescriptions for several reasons: the patients don’tlike a change in their drugs’ brand name and it is hard to explain ‘‘biosimilarity’’.Already acknowledging that some patients are reluctantly using generics, there is achance for a bigger reluctance facing biosimilars. The patients don’t have an absolutenecessity to switch if their health coverage neutralizes the costs; the prescribingdoctor must monitor the change. Therefore, competitors try to be first on the hospitalmarkets with very low prices for these products (like EPOs), for the real benefit isoften to be made in the ensuing outpatient market: (retail prices). But the penetrationof hospital market implies for biosimilars to be referenced by competent authorities,to have demonstrated the expected savings to be generated, and puts into question,beyond these products’purchase price, the real cost of their hospital use.

The Issue of Biosimilars’ Cost of Using

The cost of using is made of the addition (at purchase price) of costs supported bythe buyer and linked to administrative, logistical, and clinical protocol procedures,etc., implemented. Also, as communication strategies of competitors concerningthe risks of using are directed towards doctors, communication strategies

Biosimilars: Challenges Raised by Biosimilars 73

concerning the cost of using are directed towards pharmacists when these arepurchase-managers.

This pattern in Fig. 1 is emblematic: at the low purchase price of the biosimilar,it adds slices of cost of using. It suggests that the biosimilars’ economic benefit isneutralized by these ‘‘hidden costs.’’ This pattern was not designed for a particularproduct. It aims to invite users tempted by a biosimilar to prefer a 2nd generationinnovating drug when there is one. However it provides a good basis on which toreflect on a mapping of potential extra costs.

Is There a Non-Equivalence Cost Between Dosages for Biosimilarsand Their Reference Biomedicines?The first slice (a) of additional costs suggests that, to get the same level of effect asthe originator product, the first generation biological, a higher amount of substancehas to be used, seeming to be a general characteristic of biosimilars. Certainlycases of biosimilars’ or biological medicinal products’ lower efficacy, even inef-ficacy, have been reported in international literature; but they were all concerningproducts marketed outside the European Community, probably never assessed interm of their efficacy. A faulty equivalence is not thinkable for biosimilarsapproved in Europe, given the evaluation criteria of clinical efficacy required at thetime of MA submission (see chapter ‘‘From the biosimilar concept to the MA’’).For biosimilars approved in Europe, no case of a faulty equivalence has beenreported to EMA. Consequently, this slice has a documentary interest formedicinal products marketed on foreign markets not or poorly controlled (and in

Fig. 1 Theory of cost of using a biosimilar (according to Bols, JEAHP 2008)

74 F. Megerlin

http://dx.doi.org/10.1007/978-2-8178-0336-4_2

cases of counterfeit reference biomedicinal products), but it is not relevant withinthe European regulatory framework.

Is There a Non-Equivalence Cost Between Dosages for Biosimilarsand Their Second Generation Biomedicines?The second cost slice (b) suggests that, to get the same effect level as the secondgeneration biomedicine, a larger quantity (or an equivalent-value) of biosimilarshould be used. But when that is true (for example, epoietin vs darbopoietin), thesame comparison works between the first generation product that was used as areference for the biosimilar, and the 2nd generation product. In fact, the patternsuggests a preference for 2nd generation biomedicine over all its predecessors.One cannot make the general statement that using 2nd biomedicine is less costlythan the use of the 1st generation product: this calls for an analysis per product,and then must be compared the costs of using, not only the biosimilar’s cost ofusing, vs the purchase cost for the 2nd generation biomedicine. Besides, this wouldfirst call for a ‘‘2nd generation’’ to be available. This is not systematically the case:for instance, insulin analogs are not in themselves 2nd generation biomedicines;they allow therapeutic ‘‘new strategies,’’ but don’t replace insulin itself. Such isnot the case for erythropoietins alpha or beta which may be replaced by darbo-poietin. Ultimately, these 2nd generation products don’t necessarily mean that theprevious strategies, using the 1st generation products, have to be discarded as thetwo strategies complete each other.

Is There a Pharmacovigilance Cost Specific to Biosimilars?The fourth slice (d) additional costs represents pharmacovigilance. This slice ispresented as ‘‘difficult to quantify’’. If by pharmacovigilance one means the bodyof vigilance and notifications applicable in a permanent manner to all medicine, itscost cannot be avoided, be it biomedicine or their biosimilars (for the Eprex�pharmacovigilance case concerned an innovating product, see chapter‘‘Immunogenicity’’). If, on the contrary, by pharmacovigilance one means lessspecifically the user’s contribution to the implementation by the biosimilar’s MAholder of a ‘‘risk management plan’’, then such a specific cost is not debatable, butis it specific?

Is There a Cost for a Risk Management Plan for Biosimilars?Yes. The risk management plan is an MA component (see chapter ‘‘Immunogenicity’’).It is therefore imperative and constitutes a real additional cost (time, clinical andadministrative competence). Its significance depends on its follow-up protocol and onthe scale of its implementation in the considered hospital or group of hospitals, knowingthat the plan is implemented in the whole European Union. But what is important here isthat the additional risk management plan cost is not in the realm of biosimilars only:innovating biomedicines must also include a risk management plan in their MAapplication. The plan is also applied to 2nd generation biomedicines and is a cost ofusing, etc., that has to be added to their purchase price. Besides, this cost is temporary,


http://dx.doi.org/10.1007/978-2-8178-0336-4_3

http://dx.doi.org/10.1007/978-2-8178-0336-4_3

as it may disappear all along the risk management plan’s evolution and the mastering ofinitially identified risks. Thus, this slice of additional cost is legitimate but temporary,not specific to biosimilars, and not proportional to a product price.

Is There a Cost if Switching from a Biomedicine to its Biosimilar?The third slice of cost (c) is supposed to represent a quantifiable cost for switchingtreatments (see chapter ‘‘Substitution and interchangeability’’). The possibility of aswitch must be considered both ways: from reference biomedicine to biosimilar,and inversely, for instance in case of a break or change in the supply chain. Thatcalls for several reflections. If monitoring a possible switch is a must, switching isnot recommended because it would increase the immunogenic risk. It is not ascientific certitude but a highly plausible hypothesis that could be verified only bynon ethical trials; the new molecule is not necessarily more immunogenic in itselfthan its reference, but the switch can unbalance, through consecutive and alter-nated exposure to molecules similar but not strictly identical, tolerance to thereference molecule that the patient has gradually acquired. Therefore it is notrecommended, one way or another. Consequently, the switching cost should not beshown in the graph as systematic and permanent, as suggested by the (c) slice.What is at stake, for producers in competition, is the primo-prescription of theirproduct.

Is There a Cost Linked to a Restricted Use of Biosimilars?The 5 and 7th cost slices (e and g) generally suggest some limitations in the use,and some restrictions in the authorised indications. What about? A biomoleculemay have several treatment indications, several administration modalities (forinstance SC or IV for parenteral administration), and a more or less complexpharmacology. Thus, for a biosimilar candidate, when efficacy and/or safety datacannot be extrapolated from a studied indication to a non studied other one, or anadministration route to another, additional data are required from the developingproducer (see chapter ‘‘From the biosimilar’s concept to the MA’’). For EPO, forexample, safety (immunogenicity) data may be extrapolated from SC to IV route,but not in reverse; inversely, when safety and efficacy of an EPO’s biosimilar havebeen demonstrated in patients with chronic renal insufficiency, this biosimilar maybe used in other indications. The situation is more complex when several indi-cations are claimed for a single molecule able to interact with several receptors(anti TNF, anti-B cells, anti-VE-VF, etc.). Let’s imagine that a producer does notwant to bear the cost of additional studies because of the structure of the market hetargets, then the use of his biosimilar will be necessarily restricted only to dem-onstrated indications and modalities, duly justified of regularly extrapolated.This possible restriction must then imperatively be mentionned in the product’SPC as well as the patient information leaflet. In a health care facility, such arestriction may then impose referencing of other ‘‘similar’’ products, whoseindications or administration modalities would be, according to the assumption,differentiated. This multiple referencing may induce management costs, or

76 F. Megerlin

http://dx.doi.org/10.1007/978-2-8178-0336-4_4

http://dx.doi.org/10.1007/978-2-8178-0336-4_2

inversely scale savings, depending on clinical needs (more or less specialised),following the eventually regrouped purchase volume, and depending on the pricesobtained (the modalities of negotiation or pricing vary with the country). Somebiosimilars may be concerned, in agreement with European regulations, by arestricted use, but not in a systematic way. When these restrictions are in place,they do not necessarily induce an extra cost for the health care facility. All dependson the pharmaceutical management of purchase, a discipline that requires a sys-temic approach, a high scientific expertise and good coordination with doctors.These potential restrictions naturally underline the quality of prescription, dis-pensing, administration and traceability - as for all medicines.

Is There a Cost Linked to Logistics and Administration of Biosimilars?The sixth slice of costs (f), rightly suggests it. The new references managementplus the allocation constraints depending on the product purchase modalities, arereal potential costs but vary according to health care structures; it is artificial torepresent them in percentage of a product cost that is even not determined.Besides, the same issue exists for second generation biological medicines.

Conclusion

Although it confuses biosimilars approved in and outside the EU, this graph rightlyinvites a mapping of specific costs that increase product purchase prices. In com-munication battles, this cost is increased by biomedicine producers, and decreasedby biosimilar producers (with both camps exaggerating). In fact, the reasoningcannot settle for all these generalities: it calls for a local and rigourous approach,product by product (both originators and their biosimilars) that would involve allcost of use.

Whatever, risk management is a non negotiable priority. In what terms?

General Information on Risk Management Responsability

By contrast with scarfacing experiences reported in non european-markets (andsometimes confused in the litterature) the European framework of biosimilarsauthorisation guarantees a high level of quality and safety for these products. Tothis day, no incident has been reported about authorised biosimilars. What aboutlegal liability? The MA holder’s liability leads to technical development withoutoriginality: industrial responsibilities respond to the same rules, be the product abiologic or its biosimilar. Let’s focus on the responsibility of biosimilar users. Thissection will deal with the French approach, before dealing with the recent UScongresionnal approach.


How the Issue is Dealt with in France

Is There a Specific Legal Responsability for Biosimilars?No. producers and healthcare providers are exposed to the same legal liability asfor other medicines, in case of damages due to a faulty product or a fault.Healthcare providers are essentially liable for malpractice; they are responsible fora faulty product only if the producer is not known. In absence of a defectiveproduct or malpractice, national solidarity is activated under ‘‘risk development’’or ‘‘medical hazard,’’ to pay for possible damage. According to french law,a medicine’s undesirable effect does not mean that the product is defective, if thispotential effect is clearly mentioned with precision in the package leaflet. As aresult the producers of biomedicines and of their biosimilars provide thoroughinformation on their leaflets. In fact, warning of a risk on the SPC (Summary ofProduct Characteristics) only—to which the patient has little or no access—cannotexempt the marketing authorisation holder of his liability. All producers fall underthe permanent obligation of pharmacovigilance, health alert, possible batchwithdrawal, and leaflet updating:

Is There a Possible Specific Malpractice Associated with Biosimilars?This question is linked to another one: do health professionals have specificobligations in the context of biosimilars? Except for precautions or restrictionsspecific to marketing authorisation, the clinical use regulations fall under nationalcompetence. In France, there is no specific rule for the use of biosimilars, butbiosimilars are not listed in the generics repertoire. Subsequently, their mandatorysubstitution as for chemical generics is de facto not imposed nor recommeded (seechapter ‘‘Substitution and interchangeability’’). Moreover, some learned societieshave issued recommendations on the use of biosimilars. This type of recommen-dation affirms a high level of consensus among experts on a state of scientificknowledge. If they are not imposed because of a standard, these recommendationsare however source of a latent professionals’ obligation. Therefore their contentand scope should be studied.

Could Ignoring a Learned Society’s Recommendations be a Fault?The fact that a learned society issues recommendations does not make them anorm. But such recommendations may define means useful or even necessary forensuring care safety/efficacy, notably, for example, in cases of delay or silence ofpublic sources. Now, besides its clinical application the ‘‘state of the art’’ soobjectified may find a legal application: in case of damage that could be avoided orreduced through compliance with the recommendations, a court could considerthat a healthcare provider ignoring them had breached a latent ‘‘obligation ofmeans,’’ as this professional is supposed to update his scientific knowledge and todo his best for the patient.

78 F. Megerlin

http://dx.doi.org/10.1007/978-2-8178-0336-4_4

What is in the Recommendations as far as Biosimilars are Concerned?Again from a methodological point of view, and as an example, here are recom-mendations issued by Nephrology society, Francophone Dialysis society, andPediatric Nephrology society (2008). These recommendations are shown in theTable 1.

Let’s remember that what interests us here, is, as a procedure optimising therisk management, it is added to the state of scientific and technical knowledge.Although coming from a non-normative source, this procedure becomes virtuallyeffective for practitioners in front of the judge, either civil (private sector care) oradministrative (public sector care), hearing the case.

Are These Recommendations Exhaustive and Specific to Biosimilars?No. Recommendations are always in a temporary state of scientific knowledge andconcepts. They don’t pretend to offer a complete decision algorithm becausepracticing must integrate all normative constraints: the possible application of arisk management plan for a specified period; obligation to inform patients on thebenefit/risk ratio and on the existence (or lack) of treatment alternatives; obligationto collect patients’ informed consent on this basis, etc. It should be emphasisedthat, even if it had not been their intent, these recommendations—and rules—alsoapply to biomedicines, and not biosimilars only. The risks can perfectly beextrapolated between these two categories of products (potential variabilitybetween production sites, even between batches, change or breakdown in thesupply chain, immunogenic potential inherent to the product, the treatment dura-tion or the patient’s tolerance). This risk management should become reasonablysystematic and extend to biomedicines as well as their biosimilars. The samereasoning is valid when setting a serum bank is necessary. Let’s remember indeedthat reflecting on the risk and its generalisation was born of experience withoriginator biomedicines, and not with biosimilars (see the Eprex� case). Thatnecessarily leads to the addition of a 3rd column to the table, as follows (Table 2).

Table 1 Summary of learned recommendations

Initial prescription of a biosimilar Substitution of an innovating product by abiosimilar

It necessitates a new prescription by theauthorised doctor

A biosimilar’s traceability must be ensured atinjection time

A biosimilar’s traceability must be ensured atinjection time

An individual prescription follow-up file is putin place and updated by the prescribing doctor

An individual prescription follow-up file is putin place and updated by the prescribing doctor

Complete declaration of observed undesirableeffects

Complete declaration of observed undesirableeffects

A serum bank is created A serum bank is created


Conclusion

In France, legal liabilities are the same for dispensing and prescribing all bio-medicines, be they originator or biosimilar. But unbalanced communication on therisk inherent to any biomolecule leads to special focus on biosimilars. It underlinesthe specific application of follow-up measures, and therefore orientates theattention, even the worry of patients and practitioners alike. Yet, this unbalance incommunicating must not lead to ignoring the need of total traceability and vigi-lance whatever biomedicine (originator as well as its biosimilar) is used. The riskcan indeed be induced by possible variability between batches of the same brand,by the way of use, or by the patients profile.

The risks may affect all these products of biological origin, whatever time theirbrand has been on the market.

Brief Considerations on some international issues

A major issue has been put into light by risk management: the designation ofsubstances of biological origin, and the monitoring of treatments based on productswith different batches, production sites or brand names. After a reminder of generalreflections, the solution chosen in the US in 2010, will be presented.

Table 2 Extended summary of learned recommendations

Initial prescription of abiosimilar

Substitution of an innovatingproduct by a biosimilar

Initial prescription of abiomedicine

Substitution of an innovatingproduct by a biosimilar or thereverse! necessitates a newprescription by an authorizeddoctor

A biosimilar’s traceabilitymust be ensured at injectiontime

A biosimilar’s traceabilitymust be ensured at injectiontime

A originator biomedicinetraceability must be ensured atinjection time

An individual prescriptionfollow-up file is put in placeand updated by the prescribingdoctor



Complete declaration ofobserved undesirable effects



A serum bank is created A serum bank is created A serum bank is created

80 F. Megerlin

The General Issue of the Products’ Designation

What is the Substances’ International Non-proprietary Name (INN)?The international non-proprietary name fills in the need for a scientific nomenclatureof molecules, that identifies them in an unequivocal and neutral way on an interna-tional basis. The scientific name of the molecule is attributed by the World HealthOrganisation (WHO). It is well distinct from the commercial brand name, chosen bythe producer. When the molecule is not protected by a patent anymore, the INN maybe used by all producers competing on the market, associated with their own brandname. Using the same INN, affirming the identical scientific classification, is a keyelement to penetrate markets: it determines the doctors’ prescription, and sometimesauthorises the ‘‘automatic’’ substitution by the pharmacist (as in the UK). But theproblem resides in the fact that bearing the same INN does not mean the same thingfor chemical molecules (where molecules can be called identical), and for the field ofbiological molecules (where the molecules can only be called ‘‘similar’’).

Do Biosimilars Use Their Reference Products’ INN?It is the World Health Organisation (WHO) who attributes INNs. At the Europeanlevel, the European Medicinal products Agency will only authorise or not the productunder the INN filed by the applicant. As a result, as soon as the biosimilarity has beendemonstrated according to EU regulation (see chapter ‘‘From the biosimilar conceptto the MA’’), the biosimilar may use the same INN as the reference product, withoutany obligation to do so. That is only optional for the biosimilars’ producer, a strategicmarketing decision: some want to put the accent on the biosimilarity, and they use thesame INN (for example, several competing EPOs are marketed under the same INN‘‘EPO alpha’’). On the opposite, others wish to single out their product and apply foran MA under an INN different from the reference product’s (for example a biosimilaruses ‘‘EPO zeta’’ as it INN, when its reference product uses ‘‘EPO alpha’’ as INN).This option is not in contradiction with the biosimilars’ European regulation, insofaras the elements constitutive of the ‘‘proof of similarity,’’ in the MA biosimilars’regulation context, are not identical to those considered by WHO for allocating anINN, at the request of the considered molecule’s holder.

Is There a National Regulation on INN Use?No. As long as the centralised MA conditions are met and that the authorisation isgranted by the European agency, according to the applicant’s dossier terms, nationalauthorities are not competent for using that INN or another. But, as an identical INNmay in some countries induce a non controlled substitution, some of the Union memberstates recommend prescribing biomedicines under their brand names rather than undertheir INNs. That is the case in the UK, on the initiative of the MHRA (Medicines andHealthcare Products Regulatory Agency). In France, the issue is raised in differentterms: medicines are generally prescribed under their brand names. Besides and aboveall, biosimilars are not listed on the repertoire and their substitution, in pharmacies,is de facto not imposed (see chapter ‘‘Substitution and interchangeability’’). Thus, the


http://dx.doi.org/10.1007/978-2-8178-0336-4_2

http://dx.doi.org/10.1007/978-2-8178-0336-4_2

http://dx.doi.org/10.1007/978-2-8178-0336-4_4

fact that a reference product and its biosimilar have the same INN does not make thesubstitution automatic in France (the substitution is also not ‘‘allowed’’ for chemicalmolecules that are not listed on the generic repertoire).

Is There a Specific International Regulation of INNs for Biosimilars?Not today. The question is being discussed at the World Health Organisation level,the only authority competent in INNs matters. Several theses are in competition:producers of originator biomedicines take the position that each medicinal productof biological origin should have its own INN (that would fragment the market andmake its penetration harder). On the opposite, producers of biosimilars support theposition that their products be allowed to use the same INN as the referenceproducts (that would trivialize them and would make their diffusion on the marketeasier). The problem is that, in the first hypothesis, attributing specific and differentINNs from the reference product to each of its biosimilars (in order to recognize theorigin of the molecule) would make the INN system (International Non-proprietaryName) degenerate into an ‘‘IPN’’ system (International Proprietary Name)—theopposite of its neutral scientific classification vocation. In the second hypothesis,the granting of the same INN to a biomedicine and its biosimilars does not respondto the requirements of pharmacovigilance, that needs very precise elements oftraceability (see chapter ‘‘Substitution and interchangeability’’). Other solutionsmust be found for the classification of biological substances, which are no chemicalsubstances.

Some considerations on the US approach

Regulation of Biosimilars in the USIn 2010, the US have adopted a general regulatory framework for the approval ofbiosimilars; the guidelines for MA should be developed by the Federal DrugAdministration (an American equivalent of the European Medicine Agency). For alarge part, the US regulation has adopted the biosimilarity definition and approachelaborated by the European Union. But, while the European regulation has notentered the debate of clinical use, notably as far as substitution is concerned(see chapter ‘‘Substitution and interchangeability’’), the American regulation haveaddressed the issue, by taking a position on the qualification of what a ‘‘sameactive substance’’ means. The denomination of active substances by INN is afundamental issue in prescription, dispensation, substitution and traceability ofproducts. Yet, as stated above, the context is thorny: although the current solutionsfor biological substances denomination are not satisfying, the competence of WHOin the classification matter cannot be contested.

How did the US Circumvent the International Issue?The approach has been very clever: subject to a more precise interpretation, to begiven in guidelines that the FDA is bound to issue, the US legislation implicitlysubordinates the right, for a biosimilar, to claim its reference product’s INN, to the

82 F. Megerlin

http://dx.doi.org/10.1007/978-2-8178-0336-4_4

http://dx.doi.org/10.1007/978-2-8178-0336-4_4

fulfilment of a specific requirement. If the marketing authorisation holder wishesto claim the same INN, he must, above demonstrating the biosimilarity, demon-strate his product’s ‘‘interchangeability’’ with the reference product. It does notmean a simple additional quality of the product, but a distinct legal statute (indeedthe title of the section is ‘‘(k) Licensure of Biological Products as Biosimilar orInterchangeable’’). The interchangeability proof opens (by inference) the right tothe biosimilar’s designation by the same INN as the reference product’s, the rightof substitution by the pharmacist without intervention from the prescribing doctor,and at last the right to a minimal 12 month additional protection of data, for thefirst biosimilar of a given originator biomedicine whose interchangeability wouldbe demonstrated, (for ‘‘rewarding’’ the developer’s effort).

How does US Regulation Define Interchangeability?Interchangeability results from a scientific proof to be assessed by the FDA.According to US 2010 regulation and awaiting for FDA guidelines.

‘‘(4) SAFETY STANDARDS FOR DETERMINING INTERCHANGEABILITY.—Upon review of an application submitted under this subsection or any supplement to

such application, the Secretary shall determine the biological product to be interchange-able with the reference product if the Secretary determines that the information submittedin the application (or a supplement to such application) is sufficient to show that — ‘‘(A)the biological product — ‘‘(i) is biosimilar to the reference product; and ‘‘(ii) can beexpected to produce the same clinical result as the reference product in any given patient;and ‘‘(B) for a biological product that is administered more than once to an individual, therisk in terms of safety or diminished efficacy of alternating or switching between use ofthe biological product and the reference product is not greater than the risk of using thereference product without such alternation or switch.

This requirement is laudable in its safety intent; but as of today’s state ofscientific knowledge and available techniques, the demonstration seems to bealmost impossible, but for the less complex molecules. That makes few potentialapplicants, besides insulin for treating diabetes - a market with huge savingspotential for the US. Moreover, even if biologics were ‘‘interchangeable’’, US statemembers can decide wether they authorize the substitution or not.

For the Competing Biosimilars Involved, What are the Consequences of theLack of Interchangeability?According to US 2010 regulation, the term « interchangeable » or « interchange-ability », means that the biological product may be substituted for the referenceproduct without the intervention of the health care provider who prescribed thereference product. Under this section, the ‘‘interchangeable product’’ shall not beconsidered to have a new active ingredient. In case of non interchangeability, onecan infer that the producers of biosimilars will develop a strategy of marketpenetration equivalent to a new product’s. Competition by way of substitution isclosed, as it needs medical prescription. Besides, costs and delays required fordemonstrating interchangeability make improbable an ulterior benefit, for abiosimilars producer, to change the INN of a product he already commercialises.


That should lead to that, outside of exceptions described above, the referenceproducts and their biosimilars will for long use distinct INNs in the USA. On theother hand, the price competition could be intensified, leaning on a simplermarketing concept.

For American Health Care Professionals, What could be the Consequencesof the Lack of Interchangeability?It is doubtful that physicians will take the risk of switching originator biomedicinesby ‘‘non interchangeable’’ products, and no insurer will take the risk to push them.Consequently, patients treated before the coming of biosimilars could still be at theinnovating product’s cost. On the other hand, previously untreated patients (naive)American patients (non previously treated by biologics) will probably be treated ata biosimilar’s cost, as ‘‘bundled payments’’ of services and expensive biomedecinescould strongly incentivize the choice of biosimilars for first prescription.

Conclusion

When the US regulation makes the substitution improbable, it may, on the otherhand, intensify the price competition. Out-of-pocket payment and bundled pay-ments in the US indeed grants to the American prescribing doctor an importantsocial and economical responsibility. Throughout the world, constrained healthcarebudgets could fuel the rapid development of biosimilars markets, unless theapparition of new generation of biologics carrying higher value in health convincespayers, potentially the patient himself, and not only the prescribers.

Further Reading

EMEA Workshop on Biosimilar Monoclonal Antibodies, 2 July 2009 – Session 3Age Biotech Come To (2006) Health Affairs 25:1202–1309Schellenkens H (2009) Biosimilar therapeutics – what do we need to consider? Nephr. Diag.

Trans 2:i27–i36Pisani J, Bonduelle Y (2008) Opportunities and Barriers in the Biosimilar Mar-ket : Evolution or

Revolution for Generics Companies? PWC LondonThompson Pharma’s Red Book 2007Bols T (2008) Biosimilar in Clinical Practice – the Challenges for Hospital Phar-macists. Journal

of European Association of Hospital Pharmacists, vol 14(2):33–34Bols T, Biosimilars in Europe : from approval to clinical practice, Washington University, WDC,

sept. 2008European Generic Medicine Association Handbook for biosimilars, http://www.bogin.nl/

files/ega_biosmilarshandbook.pdfBiosimilars Series: Stakeholder Analysis A Panoramic View of the Emerging Biosimilars

Landscape (2008), DATA MONITOR DMHC 2426Cornes P (2011) The economic pressures for biosimilar drug use in cancer me-dicine. Oncologie,

Springer-Verlag. doi:10.1007/S10269-011-2017-9Megerlin F, Lopert R, Trouvin JH Biosimilars and the European experience: implications for the

United States (to be published in Health Affairs)

84 F. Megerlin

http://www.bogin.nl/files/ega_biosmilarshandbook.pdf

http://www.bogin.nl/files/ega_biosmilarshandbook.pdf

http://dx.doi.org/10.1007/S10269-011-2017-9

Afterword

Perspectives: Challenges with Biosimilars

The observant reader of the chapters included in this book will easily have realizedthat some of the assumptions and rules for biosimilars are, on a case-by-case basis,not easy to prove. For example, how far can one go with the extrapolation of databetween different clinical indications? In other words: If for a biosimilar only onekey clinical indication has been studied, can one indeed infer efficacy for anotherindication that is licensed for the reference medicinal product, without any data forthe new biosimilar? In the current regulatory approach to this issue, the efficacyand safety of the biosimilar has to be justified or, if necessary, demonstratedseparately for each of the claimed indications. This means that there needs to bedata, unless otherwise justified. This will most probably not change in future, butmany upcoming products that could emerge as biosimilars will require adiscussion on one of the central aspects, i.e. whether or not the samemechanisms of action or the same receptor(s) are involved in all indications.This can be difficult to establish for some compounds, for example those whichinterfere in complex cytokine networks, for diseases where the pathogenesis is notyet clearly defined; or for biological where the mechanism of action is not yet fullyunderstood. Related to this question is the extrapolation of safety: if we accept thatefficacy data from one indication sufficiently allow us to include that the biosimilarwill work in the other indication as well (as does the reference medicinal product),then can we be sufficiently reassured that also the safety profile would also besimilar? Can safety be extrapolated? One could here, on a case-by-case basis,perhaps employ the tool of post-marketing pharma co vigilance and study safety inseveral indications once the biosimilar is licensed. Another question remains,however: If indeed regulators would accept such extrapolations on scientificgrounds, would treating physicians also be willing to accept and use a biosimilarfor all indications, especially for more critical clinical indications like anticancermedicines? On scientific grounds this is possible; otherwise a biosimilar would notbe licensed in such indication. However, public perception appears to be differentat times. This is why for clinicians knowledge of what a biosimilar is and how it isscientifically developed is indeed key for clinicians. Related to this challenge is thequestion that was already discussed in the opening remarks on this book: If the

J.-L. Prugnaud and J.-H. Trouvin (eds.), Biosimilars,DOI: 10.1007/978-2-8178-0336-4, � Springer-Verlag France 2013

85

most sensitive model is a particular clinical indication and a particular clinicalendpoint, this may not be the most severely affected patient population, or not themost representative clinical presentation for a particular disease. If biosimilaritywere established here, would one accept the assumption that the biosimilar will beequally effective (and safe) in more challenging clinical scenarios, e.g. heavilypretreated patients, or patients with more advanced disease? This is also currentlyunder debate, and it is not yet clear what, from a scientific point of view, shouldprevail: The highest probability of establishing biosimilarity, or the best possibleway to establish that the biosimilar also works in clinically challenging scenarios,potentially at the risk that the data is less conclusive since it is significantlyconfounded by numerous factors.

Could We be too Sensitive?

Another challenge that is indeed interesting is the shift of perception of the power ofstate-of-the-art physicochemical and biological analytical methods to detectpotential differences: Whereas in the past question was ‘‘are the methodssensitive enough?’’, this question has, for some methods, now rather changed to‘‘what do differences mean?’’, since indeed some methods have reached a highdegree of sensitivity, but science is for some aspects not advanced enough to predictor estimate the clinical impact (if any) of measured differences. This will mostprobably put more burdens on the comparative non-clinical and clinicaldevelopment. But—what if, for non-clinical studies, in this scenario there is onlya relevant animal model (i.e., one in which the biosimilar is pharmacologicallyactive due to the expression of the receptor or an epitope, in the case of monoclonalantibodies) that is a non-human primate species? A powerful powered comparativetoxicology study that is designed to detect differences in toxicity may require alarge number of animals. This is not only expensive, but also potentially ethicallyquestionable. And, should one focus on on-target toxicity, i.e. toxicity related topharmacological activity? Here one would indeed have to ask for a relevant animalmodel. Or, should one study off-target toxicity?

Could one do this in a non-relevant species, for example to test impurities forany undesirable side effects? This would clearly represent a paradigm shift, sincedata from a non-relevant species is not usually required for biologicals. One wouldthen perhaps, have to focus on the clinical study, and indeed the question thencomes back to the extent of extrapolation to other clinical indications one wouldthen be ready to accept, if there were less non-clinical data.

Challenges: Not Always Scientific

Whilst challenging, many points can fortunately be addressed using solid data andsound scientific justifications. However, there are issues that are more related to the

86 Afterword

perception of a biosimilar—be it justified or not—in the scientific community. Howcertain perceptions are explained, for example that biosimilars are allegedly ‘‘lesssafe because less well tested’’, is beyond the author to discuss. Certainly, the readerof this book will now be in a position to appreciate the complexity of the issue, andhow high the standards for ‘‘true biosimilars’’ are indeed set. But, there certainly areemerging concerns at least amongst EU regulators about an inappropriate use of theterm ‘‘biosimilar’’ and its potential clinical implications. For example, at a recentconference a physician voiced his doubts about biosimilars, since they may noteven contain the active substance—here, the term biosimilar was apparently usedinappropriately, since this colleague referred to ‘‘counterfeit medicines’’ (i.e.,deliberately and fraudulently mislabeled with respect their identity and/or source),not to ‘‘biosimilar medicines’’. The fact that various different terms have emergedinternationally for ‘‘copy biopharmaceuticals’’, including ‘‘biosimilars’’ (in theEU), ‘‘follow-on biologicals’’, ‘‘subsequent-entry biologicals’’, ‘‘me toobiologicals’’, ‘‘biogenerics’’, ‘‘non-innovator’’ proteins, or ‘‘2nd generationproteins’’ is certainly not helpful. Readers of this book will however surely knowwhat the word ‘‘biosimilar’’ indeed means.

Outlook

Clearly, the further expansion of the biosimilar framework to more complexmolecules will raise numerous questions, in fact beyond the question if it is possible todevelop more complex biologicals than biosimilars. The question, for some products,will indeed be if it is at all feasible or even financially attractive. For example,complex vaccines, if developed as biosimilars, may be extremely difficult to becompared on a physicochemical and biological analytical level, and a comparativeequivalence clinical trial may have to be extremely large to be sufficiently able todetect potential differences to the reference product. The question indeed arises ifsuch a vaccine would be easier to be developed as a ‘‘standalone’’ product, even moreso if accepted surrogates for clinical efficacy like immunogenicity parameters foranti-vaccine antibodies exist. With extremely complex medicines like AdvancedTherapy Médicinal Products arising (including gène thérapies, cell based médicinalproducts, and tissue engineered products), one may face complexities that are wellbeyond the possibilities of biosimilars. But, only time will tell.

Christian K. SchneiderChairman, CHMP Working Party on Similar Biological

(Biosimilar) Médicinal Products Working Party (BMWP),European Medicines Agency, London, United Kingdom.

Paul-Ehrlich-lnstitut, Fédéral Agency for Sera and Vaccines,Langen, Germany.

Twincore Centre for Expérimental and Clinical Infection Research,Hannover, Germany.

Afterword 87

Documents

Biosimilars - University of Delhimicrobio.du.ac.in/web3/uploads/Microbiology Uploads... · Biosimilars: A Philosophy? Setting the Scene for Biosimilars Biological medicinal products